Abstract

Background

Sow removal continues to be both economically and ethically an important issue. To implement preventive measures the demographic, management and housing factors behind suboptimal culling and mortality proportions should be understood better.

Objective

This study aims to firstly, describe current management and housing factors and secondly, to investigate the associations between them and removal figures using school reports and questionnaires from two Portuguese schools.

Method

Descriptive results are generated stratified by gender, student performance groups and high alcohol usage.A multiple correspondence analysis is used to analyze social, demographic, student and parental characteristics and high alcohol usage.

Results

The result from multiple correspondence analysis shows that there is association between high alcohol usage and social and demographic features. There are individuals which share both high frequency of high alcohol consumptions and high frequencies for one or more class failures, lowest performance with regard to final grade, no willingness to higher eduction, older age, guardian being other than mother or father and going out frequently. In addition, they share a low frequency for mother´s and father´s higher eduction. Surprisingly, good health status defined by the respondent him or herself was quite frequent among the group members having high alcohol usage in common.

Conclusion

There is an indication that some sociodemographic factors have joint effects. It is important to confirm the associations using advanced techniques, e.g. by applying theory of planned behavior to study the relations among personal beliefs, attitudes, behavioral intentions and behaviour and other individual as well as parental features to investigate the risk factors for high alcohol usage.

Introduction

Data including grades, demographic, social and school related features were collected in two Portuguese schools using school reports and questionnaires and stored as two separate datasets regarding performance in distinct subjects, namely Mathematics and Portuguese. The original data of the analysis in this study are freely available as a zip file with metadata.For the purpose of this study the datasets were joined and edited according to this R script. The variables not used for joining the two data sets were combined by averaging them. Variables were further categorized to decrease the total amount of variable categories. This approach simultanously decreases the amount of information, but, on the other hand facilitates interpreation of the results. Instead of using labels for the created categories the exact ranges were used to facilitate the interpretation. In addition, a binary Alcohol_use was created by using two separate five scaled (very low-very high) variables, namely alcohol use on weekdays and during weekends. A treshold value of more than low (2) was choses for the high alcohol usage either on weekdays or on weekends.Furthermore, the variables were renamed to ease and clarify the graphical display of the results.

Structure of the dataset

Data are loaded. The final data set includes 434 respondents and 19 factorial and one supplementary, quantitative variable. The names of the variables and their explanations are listed below.

med<-read.csv(file="medyhd22mca.csv",header=TRUE)
mediso<-read.csv(file="medyhdiso.csv",header=TRUE)
tilat<-read.csv(file="3006masterfile.csv",header=TRUE)
tilat<-tilat[1:43,]

med<-med%>%mutate_all(as.factor)
#med$OUT_SOW_cullproNUM 
#colnames(med)=="OUT_SOW_cullproNUM"
#colnames(med)=="OUT_SOWmortpro"
med$OUT_SOW_cull_proNUM<-as.numeric(med$OUT_SOW_cull_proNUM)
med$OUT_SOW_mort_proNUM<-as.numeric(med$OUT_SOW_mort_proNUM)
mediso$OUT_SOW_cull_proNUM<-as.numeric(med$OUT_SOW_cull_proNUM)
mediso$OUT_SOW_mort_proNUM<-as.numeric(med$OUT_SOW_mort_proNUM)
mediso$OUT_SOW_cull_dic<-as.factor(med$OUT_SOW_cull_dic)
mediso$OUT_SOW_mort_dic<-as.factor(med$OUT_SOW_mort_dic)
medcat<-med %>% select(-ends_with("NUM_NO"),-ends_with("dosage_NO"),-ends_with("NUM"))
#medcat<-medcat%>%mutate_all(as.factor)
mednum<-med %>% select(ends_with("NUM_NO"),(ends_with("dosage_NO")),(ends_with("NUM")))
#mednum<-mednum%>%mutate_all(as.numeric)
medmca<-medcat
medmca$OUT_SOW_mort_proNUM<-as.numeric(med$OUT_SOW_mort_proNUM)
medmca$OUT_SOW_cull_proNUM<-as.numeric(med$OUT_SOW_cull_proNUM)
dfalc<-medmca
#Label<-as.matrix(colnames(dfalc))

#Variable<-c("Gender", "Age categorized",  "Parent's cohabitation status","Mother´s educational status categorized (less than secondary education, secondary education, higher education", "Father´s educational status categorized (less than secondary education, secondary education, higher education", "Mother´s job (teacher, health care, civil services, at home, other)", "Father´s job (teacher, health care, civil services, at home, other)", "Student´s guardian: mother, father, other", "Family educational support", "Willingness to take higher education", "Relationship", "Extra-cullicular activities", "Familial relationships categorized (very bad to bad, average, good to excellent)","Going out with friends categorized (very low or low, average, high or very high)","Health status categorized (very bad to bad, average, good to very good)", "Amount of failed classes: none/more than one)", "Amount of school absences one or less, 2-6hours, more than 6 hours", "Final grade categorized by quartiles", "Final grade", "Alcohol consumption more than two either during the week or at weekends")

#Level<-as.matrix(dfalc %>% sapply(levels))

#om<- data.frame(Label,Variable,Level)

#om$Level[3]<-"A(Alone),T(Together)"
#om$Level[19]<-"numeric from 0 to 20"
#rownames(om)<-NULL
#kable(om, title="Basic elements of the dataset","html") %>%
#  kable_styling(bootstrap_options = "striped", full_width = F)

Descriptive results

Firstly, a detailed variable summary is presented. Thereafter, summary tables are created using two different strata: median mortality and culling. Finally, the variables are visualized by barplots stratified by mortality and culling to capture interesting relations.

Overall summary

 

library(settings)
mediso$OUT_SOW_cull_proNUM<-as.numeric(med$OUT_SOW_cull_proNUM)
mediso$OUT_SOW_mort_proNUM<-as.numeric(med$OUT_SOW_mort_proNUM)
mediso$R_PR_sowsinsection_NUM<-tilat$R_PR_sowsinsecNUM_NO
mediso$BASIC_size_pigs_NUM<-tilat$BASIC_SIZE_PigsNUM
mediso$BASIC_size_sows_NUM<-tilat$BASIC_SIZE_Sows_NUM
mediso$BASIC_STRESSLEVEL_verymucherpal_some_NUM<-tilat$BASIC_STRESSLEVEL_verymucherpal_some
mediso$M_farNSAIDS100_NUM<-tilat$M_farNSAIDS100_NUM
mediso$M_farAB100_NUM<-tilat$M_farAB100_NUM
mediso$M_pregAB100_NUM<-tilat$M_pregAB100_NUM
mediso$MG_BR_animdirt_NUM<-tilat$MG_BR_dirt_NUM_NO
mediso$MG_BR_artinspro_NUM<- tilat$INS_artinsproNUM
mediso$R_BR_floorsolid_NUM<-tilat$R_BR_floorsolid_NUM_NO
mediso$MG_PR_animalsdirty_NUM<-tilat$MG_PR_dirt_NUM_NO
mediso$R_PR_sowgroups_NUM<-tilat$R_PR_sowsNUM_NO
mediso$R_PR_areapersow_NUM<-tilat$R_PR_areapersow_NUM_NO
mediso$R_PR_dirtypens_NUM <-tilat$R_PR_dirtNUM_NO
mediso$R_PR_floorsolid_NUM<-tilat$R_PR_floorsolidNUM_NO
mediso$MG_FAR_oxuseper10farrowings_NUM<-tilat$M_OX_10far_NUM_NO
mediso$R_FAR_pensize_NUM<-tilat$R_FAR_pensinsecNUM_NO
mediso$R_FAR_dirtypens_NUM<-tilat$MG_FAR_dirt_NUM_NO
mediso$R_FAR_floorsolid_NUM<-tilat$R_FAR_floorsolid_all0_100_100_2_muu1
mediso$MG_sows_perworker_NUM<-tilat$MG_SOWSperworkeredit_57_113_147_NUM_NO
mediso$BASIC_edulevel<-tilat$BASIC_edulevel

medcati<-mediso %>% select(-ends_with("NUM_NO"),-ends_with("dosage_NO"),-ends_with("NUM"))
medcati<-medcat%>%mutate_all(as.factor)
mednumi<-mediso %>% select(ends_with("NUM_NO"),(ends_with("dosage_NO")),(ends_with("NUM")))
mednumi<-mednum%>%mutate_all(as.numeric)

reset(options)
options("scipen"=10, "digits"=2)
tab1<-CreateTableOne(vars=colnames(mediso) ,data=mediso,factorVars = colnames(medcati))
summary(tab1)
## 
##      ### Summary of continuous variables ###
## 
## strata: Overall
##                               n miss p.miss   mean    sd median   p25 p75
## B_Biosecused                 43    0      0   0.49   0.5    0.0   0.0   1
## B_MG_R_FR_empty              43    0      0   0.60   0.5    1.0   0.0   1
## B_MG_R_FR_wash               43    0      0   0.77   0.4    1.0   1.0   1
## B_MG_R_FR_washmittel         43    0      0   0.23   0.4    0.0   0.0   0
## B_MG_R_PR_allinallout        43    0      0   0.16   0.4    0.0   0.0   0
## B_MG_R_PR_separate           43    0      0   0.70   0.5    1.0   0.0   1
## B_MG_R_PR_wash               43    0      0   0.14   0.4    0.0   0.0   0
## B_V_sirco                    43    0      0   0.30   0.5    0.0   0.0   1
## BASIC_edulevel               43    0      0   3.21   0.8    3.0   3.0   4
## BASIC_Interviewed            43    0      0   1.86   0.6    2.0   1.5   2
## M_fever                      43    0      0   0.14   0.4    0.0   0.0   0
## M_lame                       43    0      0   0.72   0.5    1.0   0.0   1
## M_rAB                        43    0      0   0.14   0.4    0.0   0.0   0
## M_rIND                       43    0      0   0.09   0.3    0.0   0.0   0
## M_secr                       43    0      0   0.12   0.3    0.0   0.0   0
## M_injury                     43    0      0   0.37   0.5    0.0   0.0   1
## MG_BR_artinspro_050_5099_100 43    0      0   1.60   0.6    2.0   1.0   2
## MG_BR_calm                   43    0      0   1.07   0.3    1.0   1.0   1
## MG_BR_feedclean              43    0      0   0.19   0.4    0.0   0.0   0
## MG_BR_nopregus               43    0      0   1.86   0.7    2.0   1.0   2
## MG_BR_ster                   43    0      0   0.14   0.4    0.0   0.0   0
## MG_PR_feedtimes              43    0      0   2.12   0.3    2.0   2.0   2
## MG_R_PR_sowsincratespostmix  43    0      0   0.23   0.4    0.0   0.0   0
## MG_FAR_feedtimes             43    0      0   2.86   0.4    3.0   3.0   3
## R_FAR_noise                  43    0      0   0.44   0.5    0.0   0.0   1
## V_ery                        43    0      0   1.02   0.2    1.0   1.0   1
## MG_sickpen_yn                43    0      0   0.77   0.4    1.0   1.0   1
## V_parvo                      43    0      0   1.02   0.2    1.0   1.0   1
## V_coli                       43    0      0   1.02   0.3    1.0   1.0   1
## V_ClC                        43    0      0   0.07   0.3    0.0   0.0   0
## V_ClA                        43    0      0   0.00   0.0    0.0   0.0   0
## V_SI                         43    0      0   0.09   0.3    0.0   0.0   0
## V_APP                        43    0      0   0.12   0.3    0.0   0.0   0
## OUT_SOW_cull_proNUM          43    0      0  12.86   7.4   12.0   6.5  18
## OUT_SOW_mort_proNUM          43    0      0  21.02  12.5   21.0  10.5  32
## R_PR_sowsinsection_NUM       43    0      0 127.05 129.0   80.0  49.0 152
## BASIC_size_pigs_NUM          43    0      0 381.51 735.8   40.0   0.0 390
## BASIC_size_sows_NUM          43    0      0 428.84 424.5  270.0 102.5 635
## M_farNSAIDS100_NUM           43    2      5  27.89  35.3   10.0   2.5  40
## M_farAB100_NUM               43    2      5   8.97  16.1    5.0   1.0  10
## M_pregAB100_NUM              43    2      5   0.85   1.6    0.5   0.0   1
## MG_BR_animdirt_NUM           43    4      9  16.67  23.3   10.0   0.0  20
## MG_BR_artinspro_NUM          43    0      0  93.79  17.4  100.0  98.0 100
## R_BR_floorsolid_NUM          43    0      0  80.14  12.0   80.0  75.0  82
## MG_PR_animalsdirty_NUM       43    0      0  24.19  20.8   20.0  10.0  30
## R_PR_sowgroups_NUM           43    0      0  16.01  18.1   10.0   7.0  17
## R_PR_areapersow_NUM          43    0      0   3.22   0.9    3.1   2.6   4
## R_PR_dirtypens_NUM           43    4      9  21.79  28.0   10.0   0.0  20
## R_PR_floorsolid_NUM          43    0      0  72.51  22.3   70.0  60.0 100
## R_FAR_pensize_NUM            43    0      0  27.95  19.1   24.0  16.0  38
## R_FAR_dirtypens_NUM          43    0      0  15.58  17.2   10.0   0.0  20
## R_FAR_floorsolid_NUM         43    0      0   0.84   0.6    1.0   0.5   1
## MG_sows_perworker_NUM        43    0      0 119.60  84.1  113.3  56.7 147
##                              min  max   skew  kurt
## B_Biosecused                   0    1  0.048 -2.10
## B_MG_R_FR_empty                0    1 -0.444 -1.89
## B_MG_R_FR_wash                 0    1 -1.312 -0.29
## B_MG_R_FR_washmittel           0    1  1.312 -0.29
## B_MG_R_PR_allinallout          0    1  1.894  1.66
## B_MG_R_PR_separate             0    1 -0.892 -1.26
## B_MG_R_PR_wash                 0    1  2.157  2.78
## B_V_sirco                      0    1  0.892 -1.26
## BASIC_edulevel                 2    5  0.750  0.51
## BASIC_Interviewed              1    3  0.053 -0.18
## M_fever                        0    1  2.157  2.78
## M_lame                         0    1 -1.021 -1.01
## M_rAB                          0    1  2.157  2.78
## M_rIND                         0    1  2.905  6.75
## M_secr                         0    1  2.481  4.36
## M_injury                       0    1  0.549 -1.78
## MG_BR_artinspro_050_5099_100   0    2 -1.185  0.50
## MG_BR_calm                     1    2  3.501 10.76
## MG_BR_feedclean                0    1  1.672  0.83
## MG_BR_nopregus                 1    3  0.173 -0.73
## MG_BR_ster                     0    1  2.157  2.78
## MG_PR_feedtimes                2    3  2.481  4.36
## MG_R_PR_sowsincratespostmix    0    1  1.312 -0.29
## MG_FAR_feedtimes               2    3 -2.157  2.78
## R_FAR_noise                    0    1  0.243 -2.04
## V_ery                          1    2  6.557 43.00
## MG_sickpen_yn                  0    1 -1.312 -0.29
## V_parvo                        1    2  6.557 43.00
## V_coli                         0    2  0.399  6.40
## V_ClC                          0    1  3.501 10.76
## V_ClA                          0    0    NaN   NaN
## V_SI                           0    1  2.905  6.75
## V_APP                          0    1  2.481  4.36
## OUT_SOW_cull_proNUM            1   27  0.185 -1.08
## OUT_SOW_mort_proNUM            1   42  0.010 -1.21
## R_PR_sowsinsection_NUM        10  600  2.062  4.43
## BASIC_size_pigs_NUM            0 3000  2.648  6.68
## BASIC_size_sows_NUM           37 2100  1.758  4.33
## M_farNSAIDS100_NUM             0  100  1.307  0.24
## M_farAB100_NUM                 0  100  4.775 26.44
## M_pregAB100_NUM                0   10  4.840 27.33
## MG_BR_animdirt_NUM             0  100  2.447  7.01
## MG_BR_artinspro_NUM           20  100 -3.816 14.61
## R_BR_floorsolid_NUM           50  100 -0.176  0.43
## MG_PR_animalsdirty_NUM         0   80  1.447  1.91
## R_PR_sowgroups_NUM             2  100  3.350 12.60
## R_PR_areapersow_NUM            1    6  0.695  0.22
## R_PR_dirtypens_NUM             0  100  1.847  2.58
## R_PR_floorsolid_NUM           20  100 -0.478 -0.13
## R_FAR_pensize_NUM              0  112  2.324  8.40
## R_FAR_dirtypens_NUM            0   80  1.583  3.49
## R_FAR_floorsolid_NUM           0    2 -0.006 -0.01
## MG_sows_perworker_NUM         11  395  1.527  3.03
## 
## =======================================================================================
## 
##      ### Summary of categorical variables ### 
## 
## strata: Overall
##                                       var  n miss p.miss
##                     MG_R_PR_sowsinsection 43    0    0.0
##                                                         
##                                                         
##                                                         
##                                   B_birds 43    0    0.0
##                                                         
##                                                         
##                                                         
##                     B_MG_R_FR_allinallout 43    0    0.0
##                                                         
##                                                         
##                          B_MG_R_FR_desinf 43    0    0.0
##                                                         
##                                                         
##                          B_MG_R_PR_desinf 43    0    0.0
##                                                         
##                                                         
##                             B_pestcontrol 43    0    0.0
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                         B_pestcontrolplan 43    0    0.0
##                                                         
##                                                         
##                               B_pestsigns 43    0    0.0
##                                                         
##                                                         
##                                 B_pets_in 43    0    0.0
##                                                         
##                                                         
##                           BASIC_size_pigs 43    0    0.0
##                                                         
##                                                         
##                                                         
##                           BASIC_size_sows 43    0    0.0
##                                                         
##                                                         
##                                                         
##      BASIC_STRESSLEVEL_verymucherpal_some 43    0    0.0
##                                                         
##                                                         
##                                BASIC_Type 43    0    0.0
##                                                         
##                                                         
##                            M_farNSAIDS100 43    0    0.0
##                                                         
##                                                         
##                               M_pregAB100 43    0    0.0
##                                                         
##                                                         
##                                     M_rOX 43    0    0.0
##                                                         
##                                                         
##                                M_farAB100 43    0    0.0
##                                                         
##                                                         
##                         M_OX_obstex_preox 43    0    0.0
##                                                         
##                                                         
##                                                         
##                           M_pregNSAIDS100 43    0    0.0
##                                                         
##                                                         
##                         MG_BR_animdirtmed 43    0    0.0
##                                                         
##                                                         
##    MG_BR_bedmatamount_no_alot_enough_some 43    0    0.0
##                                                         
##                                                         
##                                                         
##                            MG_BR_feedtype 43    0    0.0
##                                                         
##                                                         
##                                                         
##                    R_BR_floorsolid_0981_2 43    0    0.0
##                                                         
##                                                         
##                           R_BR_kuivaliete 43    0    0.0
##                                                         
##                                                         
##                             R_BR_PREGsame 43    0    0.0
##                                                         
##                                                         
##                       R_BR_sowspersection 43    0    0.0
##                                                         
##                                                         
##                                                         
##                                                         
##                         MG_PR_animdirtmed 43    0    0.0
##                                                         
##                                                         
##    MG_PR_bedmatamount_no_alot_enough_some 43    0    0.0
##                                                         
##                                                         
##                                                         
##                      MG_PR_feed_liq_solid 43    0    0.0
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                          MG_PR_kuivaliete 43    0    0.0
##                                                         
##                                                         
##             MG_PR_rootamount_no_alot_some 43    0    0.0
##                                                         
##                                                         
##                                                         
##                                MG_PR_type 43    0    0.0
##                                                         
##                                                         
##                                                         
##                                                         
##                         MG_R_PR_sowgroups 43    0    0.0
##                                                         
##                                                         
##                           R_PR_areapersow 43    0    0.0
##                                                         
##                                                         
##                                                         
##                              R_PR_dirtmed 43    0    0.0
##                                                         
##                                                         
##                    R_PR_floorsolid_0791_2 43    0    0.0
##                                                         
##                                                         
##                                 MG_FAR_ox 43    0    0.0
##                                                         
##                                                         
##                                                         
##                             R_FAR_pensize 43    0    0.0
##                                                         
##                                                         
##                          MG_FAR_bedamount 43    0    0.0
##                                                         
##                                                         
##                                                         
##                            MG_FAR_dirtmed 43    0    0.0
##                                                         
##                                                         
##           MG_FAR_ind_0no_1rout_2sometimes 43    0    0.0
##                                                         
##                                                         
##                                                         
##                      MG_FAR_nestmatamount 43    0    0.0
##                                                         
##                                                         
##                                                         
##                         MG_FAR_rootamount 43    0    0.0
##                                                         
##                                                         
##                                                         
##                                MG_FAR_toy 43    0    0.0
##                                                         
##                                                         
##      R_FAR_floorsolid_all0_100_100_2_muu1 43    0    0.0
##                                                         
##                                                         
##                                                         
##         MG_SOWS_perworkeredit_57_113_147_ 43    0    0.0
##                                                         
##                                                         
##                                                         
##                                                         
##                          OUT_SOW_cull_dic 43    0    0.0
##                                                         
##                                                         
##                          OUT_SOW_mort_dic 43    0    0.0
##                                                         
##                                                         
##  BASIC_STRESSLEVEL_verymucherpal_some_NUM 43    0    0.0
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##           MG_FAR_oxuseper10farrowings_NUM 43    0    0.0
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##                                                         
##               level freq percent cum.percent
##           (-Inf,49]   11    25.6        25.6
##          (151, Inf]   11    25.6        51.2
##            (49,151]   21    48.8       100.0
##                                             
##                  no   31    72.1        72.1
##                 no     1     2.3        74.4
##                 yes   11    25.6       100.0
##                                             
##                   0   25    58.1        58.1
##                   1   18    41.9       100.0
##                                             
##                 no    11    25.6        25.6
##                 yes   32    74.4       100.0
##                                             
##                  no   35    81.4        81.4
##                 yes    8    18.6       100.0
##                                             
##          catdogpois    1     2.3         2.3
##      catdogpoistrap    1     2.3         4.7
##  catdogpoistrapfirm    1     2.3         7.0
##             catpois   10    23.3        30.2
##        catpoisother    1     2.3        32.6
##         catpoistrap    5    11.6        44.2
##    catpoistrapother    1     2.3        46.5
##             nothing    1     2.3        48.8
##                pois   13    30.2        79.1
##            poistrap    8    18.6        97.7
##                trap    1     2.3       100.0
##                                             
##                  no   36    83.7        83.7
##                 yes    7    16.3       100.0
##                                             
##                  no   12    27.9        27.9
##                 yes   31    72.1       100.0
##                                             
##                  no   32    74.4        74.4
##                 yes   11    25.6       100.0
##                                             
##            (-Inf,0]   12    27.9        27.9
##             (0,390]   20    46.5        74.4
##          (390, Inf]   11    25.6       100.0
##                                             
##          (-Inf,102]   11    25.6        25.6
##           (102,635]   21    48.8        74.4
##          (635, Inf]   11    25.6       100.0
##                                             
##             average   29    67.4        67.4
##            verymuch   14    32.6       100.0
##                                             
##                 int   21    48.8        48.8
##              piglet   22    51.2       100.0
##                                             
##                   1   24    55.8        55.8
##                   2   19    44.2       100.0
##                                             
##                   1   28    65.1        65.1
##                   2   15    34.9       100.0
##                                             
##                   0   26    60.5        60.5
##                   1   17    39.5       100.0
##                                             
##                   1   24    55.8        55.8
##                   2   19    44.2       100.0
##                                             
##                   0   26    60.5        60.5
##                   1   16    37.2        97.7
##              noinfo    1     2.3       100.0
##                                             
##                   1   30    69.8        69.8
##                   2   13    30.2       100.0
##                                             
##                   1   25    58.1        58.1
##                   2   18    41.9       100.0
##                                             
##                ALOT    6    14.0        14.0
##                 niu   19    44.2        58.1
##                  no   18    41.9       100.0
##                                             
##             crate_L   36    83.7        83.7
##               kiosk    1     2.3        86.0
##              trough    6    14.0       100.0
##                                             
##                   1   36    83.7        83.7
##                   2    7    16.3       100.0
##                                             
##                   2   30    69.8        69.8
##                  12   13    30.2       100.0
##                                             
##                   0   31    72.1        72.1
##                   1   12    27.9       100.0
##                                             
##                 <20    3     7.0         7.0
##               20-50   11    25.6        32.6
##              50-100   24    55.8        88.4
##                 all    5    11.6       100.0
##                                             
##                   1   28    65.1        65.1
##                   2   15    34.9       100.0
##                                             
##                   0   13    30.2        30.2
##                ALOT   13    30.2        60.5
##                 NIU   17    39.5       100.0
##                                             
##                        3     7.0         7.0
##                 liq   24    55.8        62.8
##              liqsol    3     7.0        69.8
##                 sol   11    25.6        95.3
##               solid    1     2.3        97.7
##              solliq    1     2.3       100.0
##                                             
##                   2   26    60.5        60.5
##                  12   17    39.5       100.0
##                                             
##                   0    7    16.3        16.3
##                ALOT   16    37.2        53.5
##                 HIE   20    46.5       100.0
##                                             
##               loose    3     7.0         7.0
##                 pen   16    37.2        44.2
##           pen_stall    3     7.0        51.2
##          pen_stallL   21    48.8       100.0
##                                             
##           (-Inf,10]   23    53.5        53.5
##           (10, Inf]   20    46.5       100.0
##                                             
##            (-Inf,3]   20    46.5        46.5
##             (3,3.7]   12    27.9        74.4
##          (3.7, Inf]   11    25.6       100.0
##                                             
##                   1   24    55.8        55.8
##                   2   19    44.2       100.0
##                                             
##                   1   26    60.5        60.5
##                   2   17    39.5       100.0
##                                             
##            (-Inf,3]   17    39.5        39.5
##               (3,7]   18    41.9        81.4
##            (7, Inf]    8    18.6       100.0
##                                             
##         (-Inf,5.25]   31    72.1        72.1
##         (5.25, Inf]   12    27.9       100.0
##                                             
##                   0   27    62.8        62.8
##                   1    5    11.6        74.4
##                   2   11    25.6       100.0
##                                             
##                   1   24    55.8        55.8
##                   2   19    44.2       100.0
##                                             
##                   0   17    39.5        39.5
##                   1    4     9.3        48.8
##                   2   22    51.2       100.0
##                                             
##                   0   15    34.9        34.9
##                   1    4     9.3        44.2
##                   2   24    55.8       100.0
##                                             
##                   0    5    11.6        11.6
##                   1    6    14.0        25.6
##                   2   32    74.4       100.0
##                                             
##                  no   20    46.5        46.5
##                 yes   23    53.5       100.0
##                                             
##                   0   11    25.6        25.6
##                   1   28    65.1        90.7
##                   2    4     9.3       100.0
##                                             
##                   1   12    27.9        27.9
##                   2   10    23.3        51.2
##                   3   12    27.9        79.1
##                   4    9    20.9       100.0
##                                             
##                   0   22    51.2        51.2
##                   1   21    48.8       100.0
##                                             
##                   0   21    48.8        48.8
##                   1   22    51.2       100.0
##                                             
##                        0     0.0         0.0
##             average   13    30.2        30.2
##                much    8    18.6        48.8
##                some   16    37.2        86.0
##            verymuch    6    14.0       100.0
##                                             
##                        0     0.0         0.0
##                   0    1     2.3         2.3
##                   1    6    14.0        16.3
##                  10   10    23.3        39.5
##                   2    6    14.0        53.5
##                   3    6    14.0        67.4
##                   4    2     4.7        72.1
##                   5    3     7.0        79.1
##                   6    1     2.3        81.4
##                   7    2     4.7        86.0
##                   8    1     2.3        88.4
##                   9    4     9.3        97.7
##              noinfo    1     2.3       100.0
## 
reset(options)

 

 

Summary stratified by gender

 

Data variable summary stratified by GENDER
0 1 p test
n 21 22
MG_R_PR_sowsinsection (%) 0.011
(-Inf,49] 9 (42.9) 2 ( 9.1)
(151, Inf] 2 ( 9.5) 9 ( 40.9)
(49,151] 10 (47.6) 11 ( 50.0)
B_Biosecused (mean (sd)) 0.38 (0.50) 0.59 (0.50) 0.177
B_birds (%) 0.577
no 15 (71.4) 16 ( 72.7)
no 1 ( 4.8) 0 ( 0.0)
yes 5 (23.8) 6 ( 27.3)
B_MG_R_FR_allinallout = 1 (%) 9 (42.9) 9 ( 40.9) 1.000
B_MG_R_FR_desinf = yes (%) 14 (66.7) 18 ( 81.8) 0.430
B_MG_R_FR_empty (mean (sd)) 0.48 (0.51) 0.73 (0.46) 0.097
B_MG_R_FR_wash (mean (sd)) 0.71 (0.46) 0.82 (0.39) 0.432
B_MG_R_FR_washmittel (mean (sd)) 0.19 (0.40) 0.27 (0.46) 0.535
B_MG_R_PR_allinallout (mean (sd)) 0.14 (0.36) 0.18 (0.39) 0.737
B_MG_R_PR_desinf = yes (%) 5 (23.8) 3 ( 13.6) 0.642
B_MG_R_PR_separate (mean (sd)) 0.71 (0.46) 0.68 (0.48) 0.822
B_MG_R_PR_wash (mean (sd)) 0.19 (0.40) 0.09 (0.29) 0.358
B_pestcontrol (%) 0.338
catdogpois 0 ( 0.0) 1 ( 4.5)
catdogpoistrap 1 ( 4.8) 0 ( 0.0)
catdogpoistrapfirm 1 ( 4.8) 0 ( 0.0)
catpois 6 (28.6) 4 ( 18.2)
catpoisother 0 ( 0.0) 1 ( 4.5)
catpoistrap 4 (19.0) 1 ( 4.5)
catpoistrapother 1 ( 4.8) 0 ( 0.0)
nothing 0 ( 0.0) 1 ( 4.5)
pois 6 (28.6) 7 ( 31.8)
poistrap 2 ( 9.5) 6 ( 27.3)
trap 0 ( 0.0) 1 ( 4.5)
B_pestcontrolplan = yes (%) 2 ( 9.5) 5 ( 22.7) 0.448
B_pestsigns = yes (%) 16 (76.2) 15 ( 68.2) 0.806
B_pets_in = yes (%) 6 (28.6) 5 ( 22.7) 0.929
B_V_sirco (mean (sd)) 0.29 (0.46) 0.32 (0.48) 0.822
BASIC_edulevel (mean (sd)) 3.10 (0.77) 3.32 (0.84) 0.369
BASIC_Interviewed (mean (sd)) 1.90 (0.54) 1.82 (0.66) 0.642
BASIC_size_pigs (%) 0.144
(-Inf,0] 3 (14.3) 9 ( 40.9)
(0,390] 12 (57.1) 8 ( 36.4)
(390, Inf] 6 (28.6) 5 ( 22.7)
BASIC_size_sows (%) 0.034
(-Inf,102] 8 (38.1) 3 ( 13.6)
(102,635] 11 (52.4) 10 ( 45.5)
(635, Inf] 2 ( 9.5) 9 ( 40.9)
BASIC_STRESSLEVEL_verymucherpal_some = verymuch (%) 7 (33.3) 7 ( 31.8) 1.000
BASIC_Type = piglet (%) 9 (42.9) 13 ( 59.1) 0.448
M_farNSAIDS100 = 2 (%) 10 (47.6) 9 ( 40.9) 0.892
M_fever (mean (sd)) 0.10 (0.30) 0.18 (0.39) 0.425
M_lame (mean (sd)) 0.67 (0.48) 0.77 (0.43) 0.450
M_pregAB100 = 2 (%) 7 (33.3) 8 ( 36.4) 1.000
M_rAB (mean (sd)) 0.10 (0.30) 0.18 (0.39) 0.425
M_rIND (mean (sd)) 0.00 (0.00) 0.18 (0.39) 0.041
M_rOX = 1 (%) 8 (38.1) 9 ( 40.9) 1.000
M_secr (mean (sd)) 0.10 (0.30) 0.14 (0.35) 0.683
M_farAB100 = 2 (%) 10 (47.6) 9 ( 40.9) 0.892
M_injury (mean (sd)) 0.48 (0.51) 0.27 (0.46) 0.176
M_OX_obstex_preox (%) 0.613
0 13 (61.9) 13 ( 59.1)
1 8 (38.1) 8 ( 36.4)
noinfo 0 ( 0.0) 1 ( 4.5)
M_pregNSAIDS100 = 2 (%) 8 (38.1) 5 ( 22.7) 0.444
MG_BR_animdirtmed = 2 (%) 7 (33.3) 11 ( 50.0) 0.425
MG_BR_artinspro_050_5099_100 (mean (sd)) 1.52 (0.60) 1.68 (0.57) 0.381
MG_BR_bedmatamount_no_alot_enough_some (%) 0.018
ALOT 6 (28.6) 0 ( 0.0)
niu 9 (42.9) 10 ( 45.5)
no 6 (28.6) 12 ( 54.5)
MG_BR_calm (mean (sd)) 1.05 (0.22) 1.09 (0.29) 0.588
MG_BR_feedclean (mean (sd)) 0.05 (0.22) 0.32 (0.48) 0.022
MG_BR_feedtype (%) 0.352
crate_L 16 (76.2) 20 ( 90.9)
kiosk 1 ( 4.8) 0 ( 0.0)
trough 4 (19.0) 2 ( 9.1)
MG_BR_nopregus (mean (sd)) 1.76 (0.70) 1.95 (0.65) 0.356
MG_BR_ster (mean (sd)) 0.19 (0.40) 0.09 (0.29) 0.358
R_BR_floorsolid_0981_2 = 2 (%) 6 (28.6) 1 ( 4.5) 0.085
R_BR_kuivaliete = 12 (%) 9 (42.9) 4 ( 18.2) 0.153
R_BR_PREGsame = 1 (%) 8 (38.1) 4 ( 18.2) 0.265
R_BR_sowspersection (%) 0.095
<20 3 (14.3) 0 ( 0.0)
20-50 5 (23.8) 6 ( 27.3)
50-100 9 (42.9) 15 ( 68.2)
all 4 (19.0) 1 ( 4.5)
MG_PR_animdirtmed = 2 (%) 5 (23.8) 10 ( 45.5) 0.243
MG_PR_bedmatamount_no_alot_enough_some (%) 0.045
0 4 (19.0) 9 ( 40.9)
ALOT 10 (47.6) 3 ( 13.6)
NIU 7 (33.3) 10 ( 45.5)
MG_PR_feed_liq_solid (%) 0.741
1 ( 4.8) 2 ( 9.1)
liq 12 (57.1) 12 ( 54.5)
liqsol 2 ( 9.5) 1 ( 4.5)
sol 5 (23.8) 6 ( 27.3)
solid 1 ( 4.8) 0 ( 0.0)
solliq 0 ( 0.0) 1 ( 4.5)
MG_PR_feedtimes (mean (sd)) 2.14 (0.36) 2.09 (0.29) 0.606
MG_PR_kuivaliete = 12 (%) 12 (57.1) 5 ( 22.7) 0.046
MG_PR_rootamount_no_alot_some (%) 0.292
0 2 ( 9.5) 5 ( 22.7)
ALOT 10 (47.6) 6 ( 27.3)
HIE 9 (42.9) 11 ( 50.0)
MG_PR_type (%) 0.557
loose 2 ( 9.5) 1 ( 4.5)
pen 6 (28.6) 10 ( 45.5)
pen_stall 1 ( 4.8) 2 ( 9.1)
pen_stallL 12 (57.1) 9 ( 40.9)
MG_R_PR_sowgroups = (10, Inf] (%) 10 (47.6) 10 ( 45.5) 1.000
MG_R_PR_sowsincratespostmix (mean (sd)) 0.14 (0.36) 0.32 (0.48) 0.182
R_PR_areapersow (%) 0.450
(-Inf,3] 8 (38.1) 12 ( 54.5)
(3,3.7] 6 (28.6) 6 ( 27.3)
(3.7, Inf] 7 (33.3) 4 ( 18.2)
R_PR_dirtmed = 2 (%) 7 (33.3) 12 ( 54.5) 0.274
R_PR_floorsolid_0791_2 = 2 (%) 12 (57.1) 5 ( 22.7) 0.046
MG_FAR_ox (%) 0.684
(-Inf,3] 8 (38.1) 9 ( 40.9)
(3,7] 10 (47.6) 8 ( 36.4)
(7, Inf] 3 (14.3) 5 ( 22.7)
R_FAR_pensize = (5.25, Inf] (%) 10 (47.6) 2 ( 9.1) 0.013
MG_FAR_bedamount (%) 0.037
0 15 (71.4) 12 ( 54.5)
1 4 (19.0) 1 ( 4.5)
2 2 ( 9.5) 9 ( 40.9)
MG_FAR_dirtmed = 2 (%) 6 (28.6) 13 ( 59.1) 0.088
MG_FAR_feedtimes (mean (sd)) 2.76 (0.44) 2.95 (0.21) 0.071
MG_FAR_ind_0no_1rout_2sometimes (%) 0.596
0 9 (42.9) 8 ( 36.4)
1 1 ( 4.8) 3 ( 13.6)
2 11 (52.4) 11 ( 50.0)
MG_FAR_nestmatamount (%) 0.546
0 7 (33.3) 8 ( 36.4)
1 1 ( 4.8) 3 ( 13.6)
2 13 (61.9) 11 ( 50.0)
MG_FAR_rootamount (%) 0.188
0 2 ( 9.5) 3 ( 13.6)
1 5 (23.8) 1 ( 4.5)
2 14 (66.7) 18 ( 81.8)
MG_FAR_toy = yes (%) 8 (38.1) 15 ( 68.2) 0.095
R_FAR_floorsolid_all0_100_100_2_muu1 (%) 0.098
0 5 (23.8) 6 ( 27.3)
1 12 (57.1) 16 ( 72.7)
2 4 (19.0) 0 ( 0.0)
R_FAR_noise (mean (sd)) 0.48 (0.51) 0.41 (0.50) 0.667
V_ery (mean (sd)) 1.00 (0.00) 1.05 (0.21) 0.335
MG_sickpen_yn (mean (sd)) 0.81 (0.40) 0.73 (0.46) 0.535
MG_SOWS_perworkeredit_57_113_147_ (%) 0.071
1 8 (38.1) 4 ( 18.2)
2 7 (33.3) 3 ( 13.6)
3 4 (19.0) 8 ( 36.4)
4 2 ( 9.5) 7 ( 31.8)
V_parvo (mean (sd)) 1.00 (0.00) 1.05 (0.21) 0.335
V_coli (mean (sd)) 1.00 (0.32) 1.05 (0.38) 0.670
V_ClC (mean (sd)) 0.05 (0.22) 0.09 (0.29) 0.588
V_ClA (mean (sd)) 0.00 (0.00) 0.00 (0.00) NaN
V_SI (mean (sd)) 0.05 (0.22) 0.14 (0.35) 0.328
V_APP (mean (sd)) 0.10 (0.30) 0.14 (0.35) 0.683
OUT_SOW_cull_proNUM (mean (sd)) 11.71 (6.81) 13.95 (7.86) 0.325
OUT_SOW_mort_proNUM (mean (sd)) 10.05 (6.13) 31.50 (6.49) <0.001
OUT_SOW_cull_dic = 1 (%) 11 (52.4) 10 ( 45.5) 0.882
OUT_SOW_mort_dic = 1 (%) 0 ( 0.0) 22 (100.0) <0.001
R_PR_sowsinsection_NUM (mean (sd)) 72.81 (55.64) 178.82 (156.73) 0.006
BASIC_size_pigs_NUM (mean (sd)) 301.38 (459.94) 458.00 (932.04) 0.492
BASIC_size_sows_NUM (mean (sd)) 258.19 (253.75) 591.73 (492.07) 0.008
BASIC_STRESSLEVEL_verymucherpal_some_NUM (%) NaN
0 ( 0.0) 0 ( 0.0)
average 6 (28.6) 7 ( 31.8)
much 4 (19.0) 4 ( 18.2)
some 8 (38.1) 8 ( 36.4)
verymuch 3 (14.3) 3 ( 13.6)
M_farNSAIDS100_NUM (mean (sd)) 24.57 (33.79) 31.38 (37.35) 0.544
M_farAB100_NUM (mean (sd)) 10.18 (21.11) 7.71 (8.64) 0.629
M_pregAB100_NUM (mean (sd)) 0.60 (0.72) 1.11 (2.18) 0.314
MG_BR_animdirt_NUM (mean (sd)) 14.50 (23.05) 18.95 (24.01) 0.559
MG_BR_artinspro_NUM (mean (sd)) 93.57 (17.88) 94.00 (17.34) 0.937
R_BR_floorsolid_NUM (mean (sd)) 83.29 (13.29) 77.14 (9.97) 0.093
MG_PR_animalsdirty_NUM (mean (sd)) 15.95 (12.81) 32.05 (23.94) 0.009
R_PR_sowgroups_NUM (mean (sd)) 18.40 (21.37) 13.73 (14.57) 0.405
R_PR_areapersow_NUM (mean (sd)) 3.37 (0.92) 3.08 (0.88) 0.291
R_PR_dirtypens_NUM (mean (sd)) 15.26 (22.94) 28.00 (31.39) 0.158
R_PR_floorsolid_NUM (mean (sd)) 79.95 (22.13) 65.41 (20.47) 0.031
MG_FAR_oxuseper10farrowings_NUM (%) NaN
0 ( 0.0) 0 ( 0.0)
0 1 ( 4.8) 0 ( 0.0)
1 4 (19.0) 2 ( 9.1)
10 3 (14.3) 7 ( 31.8)
2 3 (14.3) 3 ( 13.6)
3 3 (14.3) 3 ( 13.6)
4 2 ( 9.5) 0 ( 0.0)
5 2 ( 9.5) 1 ( 4.5)
6 0 ( 0.0) 1 ( 4.5)
7 0 ( 0.0) 2 ( 9.1)
8 1 ( 4.8) 0 ( 0.0)
9 2 ( 9.5) 2 ( 9.1)
noinfo 0 ( 0.0) 1 ( 4.5)
R_FAR_pensize_NUM (mean (sd)) 19.33 (10.78) 36.18 (21.79) 0.003
R_FAR_dirtypens_NUM (mean (sd)) 11.43 (13.15) 19.55 (19.88) 0.124
R_FAR_floorsolid_NUM (mean (sd)) 0.95 (0.67) 0.73 (0.46) 0.203
MG_sows_perworker_NUM (mean (sd)) 84.76 (48.57) 152.85 (97.60) 0.006

 

The before assumed trend for females having a mother with low educational level is not statistically significant. However, there are significant gender-related differences between mother´s working place: females seem to have mothers staying at home more, whereas the proportion of mothers as teachers is twice that for males than females. Females indeed have more family support and they are almost unexeptionally willing to participate in higher eduction. Instead, males are more active. Females judge their health status less good (bordenline significant). And, finally, 31% of the females versus 50% of the males consume a lot of alcohol either during weekdays or weekends.

Summary stratified by final grade

 

Data variable summary stratified by FINAL GRADE
0 1 p test
n 22 21
MG_R_PR_sowsinsection (%) 0.102
(-Inf,49] 8 (36.4) 3 ( 14.3)
(151, Inf] 3 (13.6) 8 ( 38.1)
(49,151] 11 (50.0) 10 ( 47.6)
B_Biosecused (mean (sd)) 0.50 (0.51) 0.48 (0.51) 0.880
B_birds (%) 0.577
no 16 (72.7) 15 ( 71.4)
no 0 ( 0.0) 1 ( 4.8)
yes 6 (27.3) 5 ( 23.8)
B_MG_R_FR_allinallout (mean (sd)) 0.41 (0.50) 0.43 (0.51) 0.900
B_MG_R_FR_desinf = yes (%) 18 (81.8) 14 ( 66.7) 0.430
B_MG_R_FR_empty (mean (sd)) 0.59 (0.50) 0.62 (0.50) 0.855
B_MG_R_FR_wash (mean (sd)) 0.86 (0.35) 0.67 (0.48) 0.133
B_MG_R_FR_washmittel (mean (sd)) 0.14 (0.35) 0.33 (0.48) 0.133
B_MG_R_PR_allinallout (mean (sd)) 0.23 (0.43) 0.10 (0.30) 0.251
B_MG_R_PR_desinf = yes (%) 5 (22.7) 3 ( 14.3) 0.750
B_MG_R_PR_separate (mean (sd)) 0.73 (0.46) 0.67 (0.48) 0.674
B_MG_R_PR_wash (mean (sd)) 0.14 (0.35) 0.14 (0.36) 0.952
B_pestcontrol (%) 0.602
catdogpois 0 ( 0.0) 1 ( 4.8)
catdogpoistrap 0 ( 0.0) 1 ( 4.8)
catdogpoistrapfirm 0 ( 0.0) 1 ( 4.8)
catpois 6 (27.3) 4 ( 19.0)
catpoisother 1 ( 4.5) 0 ( 0.0)
catpoistrap 2 ( 9.1) 3 ( 14.3)
catpoistrapother 0 ( 0.0) 1 ( 4.8)
nothing 1 ( 4.5) 0 ( 0.0)
pois 8 (36.4) 5 ( 23.8)
poistrap 4 (18.2) 4 ( 19.0)
trap 0 ( 0.0) 1 ( 4.8)
B_pestcontrolplan = yes (%) 2 ( 9.1) 5 ( 23.8) 0.372
B_pestsigns = yes (%) 16 (72.7) 15 ( 71.4) 1.000
B_pets_in = yes (%) 4 (18.2) 7 ( 33.3) 0.430
B_V_sirco (mean (sd)) 0.27 (0.46) 0.33 (0.48) 0.674
BASIC_edulevel (mean (sd)) 3.18 (0.73) 3.24 (0.89) 0.822
BASIC_Interviewed (mean (sd)) 1.82 (0.59) 1.90 (0.62) 0.642
BASIC_size_pigs (%) 0.345
(-Inf,0] 8 (36.4) 4 ( 19.0)
(0,390] 10 (45.5) 10 ( 47.6)
(390, Inf] 4 (18.2) 7 ( 33.3)
BASIC_size_sows (%) 0.210
(-Inf,102] 8 (36.4) 3 ( 14.3)
(102,635] 10 (45.5) 11 ( 52.4)
(635, Inf] 4 (18.2) 7 ( 33.3)
BASIC_STRESSLEVEL_verymucherpal_some = verymuch (%) 5 (22.7) 9 ( 42.9) 0.279
BASIC_Type = piglet (%) 11 (50.0) 11 ( 52.4) 1.000
M_farNSAIDS100 (mean (sd)) 1.32 (0.48) 1.57 (0.51) 0.099
M_fever (mean (sd)) 0.18 (0.39) 0.10 (0.30) 0.425
M_lame (mean (sd)) 0.73 (0.46) 0.71 (0.46) 0.927
M_pregAB100 (mean (sd)) 1.32 (0.48) 1.38 (0.50) 0.675
M_rAB (mean (sd)) 0.14 (0.35) 0.14 (0.36) 0.952
M_rIND (mean (sd)) 0.05 (0.21) 0.14 (0.36) 0.283
M_rOX (mean (sd)) 0.32 (0.48) 0.48 (0.51) 0.301
M_secr (mean (sd)) 0.09 (0.29) 0.14 (0.36) 0.606
M_farAB100 (mean (sd)) 1.27 (0.46) 1.62 (0.50) 0.022
M_injury (mean (sd)) 0.36 (0.49) 0.38 (0.50) 0.909
M_OX_obstex_preox (%) 0.541
0 13 (59.1) 13 ( 61.9)
1 9 (40.9) 7 ( 33.3)
noinfo 0 ( 0.0) 1 ( 4.8)
M_pregNSAIDS100 (mean (sd)) 1.32 (0.48) 1.29 (0.46) 0.822
MG_BR_animdirtmed (mean (sd)) 1.36 (0.49) 1.48 (0.51) 0.466
MG_BR_artinspro_050_5099_100 (mean (sd)) 1.55 (0.60) 1.67 (0.58) 0.502
MG_BR_bedmatamount_no_alot_enough_some (%) 0.336
ALOT 2 ( 9.1) 4 ( 19.0)
niu 12 (54.5) 7 ( 33.3)
no 8 (36.4) 10 ( 47.6)
MG_BR_calm (mean (sd)) 1.09 (0.29) 1.05 (0.22) 0.588
MG_BR_feedclean (mean (sd)) 0.18 (0.39) 0.19 (0.40) 0.944
MG_BR_feedtype (%) 0.129
crate_L 16 (72.7) 20 ( 95.2)
kiosk 1 ( 4.5) 0 ( 0.0)
trough 5 (22.7) 1 ( 4.8)
MG_BR_nopregus (mean (sd)) 1.73 (0.55) 2.00 (0.77) 0.189
MG_BR_ster (mean (sd)) 0.09 (0.29) 0.19 (0.40) 0.358
R_BR_floorsolid_0981_2 (mean (sd)) 1.14 (0.35) 1.19 (0.40) 0.641
R_BR_kuivaliete (mean (sd)) 5.18 (4.77) 4.86 (4.63) 0.822
R_BR_PREGsame (mean (sd)) 0.32 (0.48) 0.24 (0.44) 0.569
R_BR_sowspersection (%) 0.532
<20 1 ( 4.5) 2 ( 9.5)
20-50 5 (22.7) 6 ( 28.6)
50-100 12 (54.5) 12 ( 57.1)
all 4 (18.2) 1 ( 4.8)
MG_PR_animdirtmed (mean (sd)) 1.27 (0.46) 1.43 (0.51) 0.295
MG_PR_bedmatamount_no_alot_enough_some (%) 0.719
0 6 (27.3) 7 ( 33.3)
ALOT 6 (27.3) 7 ( 33.3)
NIU 10 (45.5) 7 ( 33.3)
MG_PR_feed_liq_solid (%) 0.629
1 ( 4.5) 2 ( 9.5)
liq 12 (54.5) 12 ( 57.1)
liqsol 1 ( 4.5) 2 ( 9.5)
sol 7 (31.8) 4 ( 19.0)
solid 0 ( 0.0) 1 ( 4.8)
solliq 1 ( 4.5) 0 ( 0.0)
MG_PR_feedtimes (mean (sd)) 2.09 (0.29) 2.14 (0.36) 0.606
MG_PR_kuivaliete (mean (sd)) 6.55 (5.10) 5.33 (4.83) 0.429
MG_PR_rootamount_no_alot_some (%) 0.356
0 2 ( 9.1) 5 ( 23.8)
ALOT 8 (36.4) 8 ( 38.1)
HIE 12 (54.5) 8 ( 38.1)
MG_PR_type (%) 0.815
loose 1 ( 4.5) 2 ( 9.5)
pen 9 (40.9) 7 ( 33.3)
pen_stall 2 ( 9.1) 1 ( 4.8)
pen_stallL 10 (45.5) 11 ( 52.4)
MG_R_PR_sowgroups = (10, Inf] (%) 8 (36.4) 12 ( 57.1) 0.289
MG_R_PR_sowsincratespostmix (mean (sd)) 0.27 (0.46) 0.19 (0.40) 0.535
R_PR_areapersow (%) 0.136
(-Inf,3] 9 (40.9) 11 ( 52.4)
(3,3.7] 9 (40.9) 3 ( 14.3)
(3.7, Inf] 4 (18.2) 7 ( 33.3)
R_PR_dirtmed (mean (sd)) 1.41 (0.50) 1.48 (0.51) 0.667
R_PR_floorsolid_0791_2 (mean (sd)) 1.45 (0.51) 1.33 (0.48) 0.429
MG_FAR_ox (%) 0.765
(-Inf,3] 8 (36.4) 9 ( 42.9)
(3,7] 9 (40.9) 9 ( 42.9)
(7, Inf] 5 (22.7) 3 ( 14.3)
R_FAR_pensize = (5.25, Inf] (%) 6 (27.3) 6 ( 28.6) 1.000
MG_FAR_bedamount (mean (sd)) 0.73 (0.88) 0.52 (0.87) 0.452
MG_FAR_dirtmed (mean (sd)) 1.41 (0.50) 1.48 (0.51) 0.667
MG_FAR_feedtimes (mean (sd)) 2.77 (0.43) 2.95 (0.22) 0.093
MG_FAR_ind_0no_1rout_2sometimes (mean (sd)) 0.95 (1.00) 1.29 (0.90) 0.261
MG_FAR_nestmatamount (mean (sd)) 1.45 (0.80) 0.95 (1.02) 0.080
MG_FAR_rootamount (mean (sd)) 1.73 (0.55) 1.52 (0.81) 0.340
MG_FAR_toy = yes (%) 10 (45.5) 13 ( 61.9) 0.438
R_FAR_floorsolid_all0_100_100_2_muu1 (mean (sd)) 0.95 (0.58) 0.71 (0.56) 0.173
R_FAR_noise (mean (sd)) 0.55 (0.51) 0.33 (0.48) 0.169
V_ery (mean (sd)) 1.05 (0.21) 1.00 (0.00) 0.335
MG_sickpen_yn (mean (sd)) 0.82 (0.39) 0.71 (0.46) 0.432
MG_SOWS_perworkeredit_57_113_147_ (mean (sd)) 2.05 (1.00) 2.81 (1.12) 0.023
V_parvo (mean (sd)) 1.05 (0.21) 1.00 (0.00) 0.335
V_coli (mean (sd)) 1.09 (0.43) 0.95 (0.22) 0.190
V_ClC (mean (sd)) 0.05 (0.21) 0.10 (0.30) 0.533
V_ClA (mean (sd)) 0.00 (0.00) 0.00 (0.00) NaN
V_SI (mean (sd)) 0.05 (0.21) 0.14 (0.36) 0.283
V_APP (mean (sd)) 0.14 (0.35) 0.10 (0.30) 0.683
OUT_SOW_cull_proNUM (mean (sd)) 7.18 (3.57) 18.81 (5.29) <0.001
OUT_SOW_mort_proNUM (mean (sd)) 21.05 (13.49) 21.00 (11.74) 0.991
OUT_SOW_cull_dic = 1 (%) 0 ( 0.0) 21 (100.0) <0.001
OUT_SOW_mort_dic = 1 (%) 12 (54.5) 10 ( 47.6) 0.882
R_PR_sowsinsection_NUM (mean (sd)) 79.14 (65.74) 177.24 (158.80) 0.011
BASIC_size_pigs_NUM (mean (sd)) 184.50 (403.44) 587.90 (937.30) 0.072
BASIC_size_sows_NUM (mean (sd)) 322.95 (314.52) 539.76 (499.27) 0.094
BASIC_STRESSLEVEL_verymucherpal_some_NUM (%) NaN
0 ( 0.0) 0 ( 0.0)
average 9 (40.9) 4 ( 19.0)
much 1 ( 4.5) 7 ( 33.3)
some 8 (36.4) 8 ( 38.1)
verymuch 4 (18.2) 2 ( 9.5)
M_farNSAIDS100_NUM (mean (sd)) 24.79 (38.21) 31.15 (32.59) 0.570
M_farAB100_NUM (mean (sd)) 9.27 (21.88) 8.66 (6.55) 0.906
M_pregAB100_NUM (mean (sd)) 0.60 (0.68) 1.10 (2.20) 0.333
MG_BR_animdirt_NUM (mean (sd)) 12.63 (11.95) 20.50 (30.34) 0.298
MG_BR_artinspro_NUM (mean (sd)) 94.18 (17.28) 93.38 (17.94) 0.882
R_BR_floorsolid_NUM (mean (sd)) 80.91 (11.92) 79.33 (12.29) 0.672
MG_PR_animalsdirty_NUM (mean (sd)) 21.82 (17.63) 26.67 (23.79) 0.451
R_PR_sowgroups_NUM (mean (sd)) 12.75 (9.36) 19.43 (23.99) 0.232
R_PR_areapersow_NUM (mean (sd)) 3.25 (0.84) 3.19 (0.98) 0.834
R_PR_dirtypens_NUM (mean (sd)) 19.50 (28.92) 24.21 (27.55) 0.606
R_PR_floorsolid_NUM (mean (sd)) 74.09 (21.73) 70.86 (23.28) 0.640
MG_FAR_oxuseper10farrowings_NUM (%) NaN
0 ( 0.0) 0 ( 0.0)
0 1 ( 4.5) 0 ( 0.0)
1 3 (13.6) 3 ( 14.3)
10 4 (18.2) 6 ( 28.6)
2 3 (13.6) 3 ( 14.3)
3 3 (13.6) 3 ( 14.3)
4 1 ( 4.5) 1 ( 4.8)
5 2 ( 9.1) 1 ( 4.8)
6 1 ( 4.5) 0 ( 0.0)
7 1 ( 4.5) 1 ( 4.8)
8 0 ( 0.0) 1 ( 4.8)
9 3 (13.6) 1 ( 4.8)
noinfo 0 ( 0.0) 1 ( 4.8)
R_FAR_pensize_NUM (mean (sd)) 24.05 (10.30) 32.05 (24.93) 0.173
R_FAR_dirtypens_NUM (mean (sd)) 12.27 (12.70) 19.05 (20.71) 0.201
R_FAR_floorsolid_NUM (mean (sd)) 0.95 (0.58) 0.71 (0.56) 0.173
MG_sows_perworker_NUM (mean (sd)) 87.73 (50.19) 152.99 (99.59) 0.009

 

Graphical overview

Barplots by the median mortality

Red > median mortality Green < median mortality

#lets plot  
#density plots for numerical variables7
medcati<-mediso %>% select(-ends_with("NUM_NO"),-ends_with("dosage_NO"),-ends_with("NUM"))
medcati<-medcati%>%mutate_all(as.factor)
mednumi<-mediso %>% select(ends_with("NUM_NO"),(ends_with("dosage_NO")),(ends_with("NUM")))
mednumi<-mednum%>%mutate_all(as.numeric)

colNames <- names(medcati)
for(i in colNames){
    plt<-ggplot(mediso, aes_string(x=i)) + 
      geom_bar(aes(fill = OUT_SOW_mort_dic), position = "dodge", stat="count")+
      scale_fill_manual(values = c("green","red"))
      plt + guides(fill=FALSE)
      print(plt+guides(fill=F))
}

  #### Barplots by the median culling

Orange > median culling Green < median culling

#lets plot  
#density plots for numerical variables7

colNames <- names(medcati)
for(i in colNames){
    plt<-ggplot(mediso, aes_string(x=i)) + 
      geom_bar(aes(fill = OUT_SOW_mort_dic), position = "dodge", stat="count")+
      scale_fill_manual(values = c("green","orange"))
      plt + guides(fill=FALSE)
      print(plt+guides(fill=F))
}

 

Most of the mother´s working places are defined as “other” (36%) and father’s as well (36%). Altogether 14% are stay at home mother´s and 4% of the father´s are at home. There are 16% of the mothers teaching and 8% of the fathers.Especially for the females the guardian is the mother, for every fourth it is, yet, the father and for 4% another person. There is family support for 62% of the students, and it seems to be more common for females.

 

Almost everyone, 95% has positive attitude towards higher eduction. Again, it seems even more common for females. In romantic relationship are 32% of the respondents. Females seem to be a little less active than males: altogether 47% have no extracurricular activities. Family relationships are mainly described as good or very good (76%), which is expected these data coming from a Mediterranian country with high family values.

 

Altogether 35% are going out frequently, 33% not often but not rarely, and 32% quite rarely. Health status is defined as very bad or bad as often as by almost every fourth respondent. Very good or good health status is very common, though (55%). A little more than every tenth student has one or more class failures whereas 87% have none. School absences are within 6 or less hours in 77% of the cases.

 

Performance groups represent approximately the lowest, medium low, middle high and highest groups. Very low or low alcohol consumers both at weekdays and during weekends represent 60% of the respondents.However, there seems to be an expected trend towards males drinking more.

Data variable summary stratified by FINAL GRADE
0 1 p test
n 22 21
MG_SOWS_perworkeredit_57_113_147_ (%) 0.022
(-Inf,57] 9 (40.9) 3 ( 14.3)
(147, Inf] 2 ( 9.1) 9 ( 42.9)
(57,147] 11 (50.0) 9 ( 42.9)
MG_R_PR_sowsinsection (%) 0.102
(-Inf,49] 8 (36.4) 3 ( 14.3)
(151, Inf] 3 (13.6) 8 ( 38.1)
(49,151] 11 (50.0) 10 ( 47.6)
B_MG_R_FR_allinallout = 1 (%) 9 (40.9) 9 ( 42.9) 1.000
B_pestcontrolplan = yes (%) 2 ( 9.1) 5 ( 23.8) 0.372
BASIC_size_sows (%) 0.210
(-Inf,102] 8 (36.4) 3 ( 14.3)
(102,635] 10 (45.5) 11 ( 52.4)
(635, Inf] 4 (18.2) 7 ( 33.3)
M_farNSAIDS100 = 2 (%) 7 (31.8) 12 ( 57.1) 0.172
M_pregAB100 = 2 (%) 7 (31.8) 8 ( 38.1) 0.911
M_rOX = 1 (%) 7 (31.8) 10 ( 47.6) 0.455
M_farAB100 = 2 (%) 6 (27.3) 13 ( 61.9) 0.048
M_pregNSAIDS100 = 2 (%) 7 (31.8) 6 ( 28.6) 1.000
MG_BR_animdirtmed = 2 (%) 8 (36.4) 10 ( 47.6) 0.661
MG_BR_bedmatamount_no_alot_enough_some (%) 0.336
ALOT 2 ( 9.1) 4 ( 19.0)
niu 12 (54.5) 7 ( 33.3)
no 8 (36.4) 10 ( 47.6)
MG_BR_feedtype (%) 0.129
crate_L 16 (72.7) 20 ( 95.2)
kiosk 1 ( 4.5) 0 ( 0.0)
trough 5 (22.7) 1 ( 4.8)
R_BR_floorsolid_0981_2 = 2 (%) 3 (13.6) 4 ( 19.0) 0.946
R_BR_kuivaliete = 12 (%) 7 (31.8) 6 ( 28.6) 1.000
R_BR_PREGsame = 1 (%) 7 (31.8) 5 ( 23.8) 0.806
R_BR_sowspersection (%) 0.532
<20 1 ( 4.5) 2 ( 9.5)
20-50 5 (22.7) 6 ( 28.6)
50-100 12 (54.5) 12 ( 57.1)
all 4 (18.2) 1 ( 4.8)
MG_PR_animdirtmed = 2 (%) 6 (27.3) 9 ( 42.9) 0.452
MG_PR_bedmatamount_no_alot_enough_some (%) 0.719
0 6 (27.3) 7 ( 33.3)
ALOT 6 (27.3) 7 ( 33.3)
NIU 10 (45.5) 7 ( 33.3)
MG_PR_kuivaliete = 12 (%) 10 (45.5) 7 ( 33.3) 0.617
MG_PR_rootamount_no_alot_some (%) 0.356
0 2 ( 9.1) 5 ( 23.8)
ALOT 8 (36.4) 8 ( 38.1)
HIE 12 (54.5) 8 ( 38.1)
MG_PR_type (%) 0.815
loose 1 ( 4.5) 2 ( 9.5)
pen 9 (40.9) 7 ( 33.3)
pen_stall 2 ( 9.1) 1 ( 4.8)
pen_stallL 10 (45.5) 11 ( 52.4)
R_PR_dirtmed = 2 (%) 9 (40.9) 10 ( 47.6) 0.892
R_PR_floorsolid_0791_2 = 2 (%) 10 (45.5) 7 ( 33.3) 0.617
MG_FAR_ox (%) 0.765
(-Inf,3] 8 (36.4) 9 ( 42.9)
(3,7] 9 (40.9) 9 ( 42.9)
(7, Inf] 5 (22.7) 3 ( 14.3)
MG_FAR_bedamount (%) 0.333
0 12 (54.5) 15 ( 71.4)
1 4 (18.2) 1 ( 4.8)
2 6 (27.3) 5 ( 23.8)
MG_FAR_dirtmed = 2 (%) 9 (40.9) 10 ( 47.6) 0.892
MG_FAR_nestmatamount (%) 0.019
0 4 (18.2) 11 ( 52.4)
1 4 (18.2) 0 ( 0.0)
2 14 (63.6) 10 ( 47.6)
MG_FAR_rootamount (%) 0.277
0 1 ( 4.5) 4 ( 19.0)
1 4 (18.2) 2 ( 9.5)
2 17 (77.3) 15 ( 71.4)
R_FAR_floorsolid_all0_100_100_2_muu1 (%) 0.379
0 4 (18.2) 7 ( 33.3)
1 15 (68.2) 13 ( 61.9)
2 3 (13.6) 1 ( 4.8)
MG_FAR_ind_0no_1rout_2sometimes (%) 0.268
0 11 (50.0) 6 ( 28.6)
1 1 ( 4.5) 3 ( 14.3)
2 10 (45.5) 12 ( 57.1)
OUT_SOW_mort_dic = 1 (%) 12 (54.5) 10 ( 47.6) 0.882
OUT_SOW_cull_dic = 1 (%) 0 ( 0.0) 21 (100.0) <0.001
OUT_mort15 = 1 (%) 5 (22.7) 3 ( 14.3) 0.750
OUT_mort5 = 1 (%) 14 (63.6) 16 ( 76.2) 0.573
OUT_cull50 = 1 (%) 0 ( 0.0) 5 ( 23.8) 0.050
OUT_cull30 = 1 (%) 9 (40.9) 20 ( 95.2) 0.001
OUT_SOW_mort_proNUM (mean (sd)) 21.05 (13.49) 21.00 (11.74) 0.991
OUT_SOW_cull_proNUM (mean (sd)) 7.18 (3.57) 18.81 (5.29) <0.001

 

Data variable summary stratified by FINAL GRADE
0 1 p test
n 22 21
MG_SOWS_perworkeredit_57_113_147_ (%) 0.022
(-Inf,57] 9 (40.9) 3 ( 14.3)
(147, Inf] 2 ( 9.1) 9 ( 42.9)
(57,147] 11 (50.0) 9 ( 42.9)
MG_R_PR_sowsinsection (%) 0.102
(-Inf,49] 8 (36.4) 3 ( 14.3)
(151, Inf] 3 (13.6) 8 ( 38.1)
(49,151] 11 (50.0) 10 ( 47.6)
B_MG_R_FR_allinallout = 1 (%) 9 (40.9) 9 ( 42.9) 1.000
B_pestcontrolplan = yes (%) 2 ( 9.1) 5 ( 23.8) 0.372
BASIC_size_sows (%) 0.210
(-Inf,102] 8 (36.4) 3 ( 14.3)
(102,635] 10 (45.5) 11 ( 52.4)
(635, Inf] 4 (18.2) 7 ( 33.3)
M_farNSAIDS100 = 2 (%) 7 (31.8) 12 ( 57.1) 0.172
M_pregAB100 = 2 (%) 7 (31.8) 8 ( 38.1) 0.911
M_rOX = 1 (%) 7 (31.8) 10 ( 47.6) 0.455
M_farAB100 = 2 (%) 6 (27.3) 13 ( 61.9) 0.048
M_pregNSAIDS100 = 2 (%) 7 (31.8) 6 ( 28.6) 1.000
MG_BR_animdirtmed = 2 (%) 8 (36.4) 10 ( 47.6) 0.661
MG_BR_bedmatamount_no_alot_enough_some (%) 0.336
ALOT 2 ( 9.1) 4 ( 19.0)
niu 12 (54.5) 7 ( 33.3)
no 8 (36.4) 10 ( 47.6)
MG_BR_feedtype (%) 0.129
crate_L 16 (72.7) 20 ( 95.2)
kiosk 1 ( 4.5) 0 ( 0.0)
trough 5 (22.7) 1 ( 4.8)
R_BR_floorsolid_0981_2 = 2 (%) 3 (13.6) 4 ( 19.0) 0.946
R_BR_kuivaliete = 12 (%) 7 (31.8) 6 ( 28.6) 1.000
R_BR_PREGsame = 1 (%) 7 (31.8) 5 ( 23.8) 0.806
R_BR_sowspersection (%) 0.532
<20 1 ( 4.5) 2 ( 9.5)
20-50 5 (22.7) 6 ( 28.6)
50-100 12 (54.5) 12 ( 57.1)
all 4 (18.2) 1 ( 4.8)
MG_PR_animdirtmed = 2 (%) 6 (27.3) 9 ( 42.9) 0.452
MG_PR_bedmatamount_no_alot_enough_some (%) 0.719
0 6 (27.3) 7 ( 33.3)
ALOT 6 (27.3) 7 ( 33.3)
NIU 10 (45.5) 7 ( 33.3)
MG_PR_kuivaliete = 12 (%) 10 (45.5) 7 ( 33.3) 0.617
MG_PR_rootamount_no_alot_some (%) 0.356
0 2 ( 9.1) 5 ( 23.8)
ALOT 8 (36.4) 8 ( 38.1)
HIE 12 (54.5) 8 ( 38.1)
MG_PR_type (%) 0.815
loose 1 ( 4.5) 2 ( 9.5)
pen 9 (40.9) 7 ( 33.3)
pen_stall 2 ( 9.1) 1 ( 4.8)
pen_stallL 10 (45.5) 11 ( 52.4)
R_PR_dirtmed = 2 (%) 9 (40.9) 10 ( 47.6) 0.892
R_PR_floorsolid_0791_2 = 2 (%) 10 (45.5) 7 ( 33.3) 0.617
MG_FAR_ox (%) 0.765
(-Inf,3] 8 (36.4) 9 ( 42.9)
(3,7] 9 (40.9) 9 ( 42.9)
(7, Inf] 5 (22.7) 3 ( 14.3)
MG_FAR_bedamount (%) 0.333
0 12 (54.5) 15 ( 71.4)
1 4 (18.2) 1 ( 4.8)
2 6 (27.3) 5 ( 23.8)
MG_FAR_dirtmed = 2 (%) 9 (40.9) 10 ( 47.6) 0.892
MG_FAR_nestmatamount (%) 0.019
0 4 (18.2) 11 ( 52.4)
1 4 (18.2) 0 ( 0.0)
2 14 (63.6) 10 ( 47.6)
MG_FAR_rootamount (%) 0.277
0 1 ( 4.5) 4 ( 19.0)
1 4 (18.2) 2 ( 9.5)
2 17 (77.3) 15 ( 71.4)
R_FAR_floorsolid_all0_100_100_2_muu1 (%) 0.379
0 4 (18.2) 7 ( 33.3)
1 15 (68.2) 13 ( 61.9)
2 3 (13.6) 1 ( 4.8)
MG_FAR_ind_0no_1rout_2sometimes (%) 0.268
0 11 (50.0) 6 ( 28.6)
1 1 ( 4.5) 3 ( 14.3)
2 10 (45.5) 12 ( 57.1)
OUT_SOW_mort_dic = 1 (%) 12 (54.5) 10 ( 47.6) 0.882
OUT_SOW_cull_dic = 1 (%) 0 ( 0.0) 21 (100.0) <0.001
OUT_mort15 = 1 (%) 5 (22.7) 3 ( 14.3) 0.750
OUT_mort5 = 1 (%) 14 (63.6) 16 ( 76.2) 0.573
OUT_cull50 = 1 (%) 0 ( 0.0) 5 ( 23.8) 0.050
OUT_cull30 = 1 (%) 9 (40.9) 20 ( 95.2) 0.001
OUT_SOW_mort_proNUM (mean (sd)) 21.05 (13.49) 21.00 (11.74) 0.991
OUT_SOW_cull_proNUM (mean (sd)) 7.18 (3.57) 18.81 (5.29) <0.001

 

The same overviews for the smaller dataset

#density plots for numerical variables7

colNames <- names(dfalc[,1:37])
for(i in colNames){
    plt<-ggplot(dfalc, aes_string(x=i)) + 
      geom_bar(aes(fill = OUT_SOW_cull_dic), position = "dodge", stat="count")+
      scale_fill_manual(values = c("green","red"))
      plt + guides(fill=FALSE)
      print(plt+guides(fill=F))
}

#lets plot  
#density plots for numerical variables7

colNames <- names(dfalc[,1:37])
for(i in colNames){
    plt<-ggplot(dfalc, aes_string(x=i)) + 
      geom_bar(aes(fill = OUT_SOW_mort_dic), position = "dodge", stat="count")+
      scale_fill_manual(values = c("green","orange"))
      print(plt)
}

 

There are differences among the age groups in final grades. Younger student get better scores. Mother´s educational background seem to affect the grades, whereas father´s education is not that influental. The same is seen with mother´s and father´s working place. Guardian does not affect school performance, but willingness to take higher education definitely does. Surprisingly, going out with friends does not affect final grades, but health status does. There are a lot of students in a very good or good condition performing below the average. Class failures understandably worsen the grades. Low alcohol usage is less common in the best performance group.

Summary stratified by alcohol consumption

 

 

Males drink more than females, as do the younger ones as well.Father´s job affects alcohol consumption, but mother´s job, student´s guardian or family relationships seem not to. Going out is associated with alcohol usage, but health group seems not to. Class failures are more common among the heavy alcohol users, as well as the highest numbers of school absences. Mean final grade differs significantly between the groups (mean(sd)): 11.74(3.43) versus 11.06(3.04).  

Multiple correspondence analysis

Methodology

Simple form of analysing categorical data is cross-tabulation. Correspondence analysis is an extention of contingency table data and a generalization of principal component approach. Multiple correspondence analysis (MCA) is an extension of correspondence analysis and allows to investigate the pattern of relationships of several categorical dependent variables simultaneously. Applying multiple correspondence analysis helps to reduce the interaction parameters. Using the results of a MCA, it is possible to describe the structure of all the categorical variables included. The computational graphical representations covers basically every bit of information in the data by mapping each variable/individual of analysis as a point in a low-dimensional space.

To encompass, MCA has several features that distinguish it from other techniques of data analysis. It simplifies large and complex data and provides a detailed description of practically every bit of information in the data, yielding a simple, yet exhaustive scrutiny of relationships occuring by multiple pair wise comparisons. Graphically, dual displays are produced to facilitate interpretation.

Basically, the first step is that a crosstabulated frequency table is standardized to yield relative frequencies across the cells to sum up to 1.0. The aim of a MCA analysis is to represent the entries in the table of relative frequencies in terms of the distances between individual rows and/or columns in a low-dimensional space.

Analysis

MCA() function that come in the package “FactoMineR” by Francois Husson, Julie Josse, Sebastien Le, and Jeremy Mazet. Additionally, package “factoextra” is used to beautifully visualize multiple correspondence analysis.

dfalc<-med
res.mca = MCA(dfalc,quanti.sup=(38:39),quali.sup=(32:37), graph = FALSE)

There are 434 individuals and 55 variable categories. Additionally, there is one quantitative variable, which is considered illustrative.

Description of MCA output

The output of the MCA() function is a list including :

res_mca <- MCA(dfalc,quanti.sup=(38:39),quali.sup=(32:37), graph = FALSE)
print(res_mca)
## **Results of the Multiple Correspondence Analysis (MCA)**
## The analysis was performed on 43 individuals, described by 39 variables
## *The results are available in the following objects:
## 
##    name               
## 1  "$eig"             
## 2  "$var"             
## 3  "$var$coord"       
## 4  "$var$cos2"        
## 5  "$var$contrib"     
## 6  "$var$v.test"      
## 7  "$ind"             
## 8  "$ind$coord"       
## 9  "$ind$cos2"        
## 10 "$ind$contrib"     
## 11 "$quanti.sup"      
## 12 "$quanti.sup$coord"
## 13 "$quali.sup"       
## 14 "$quali.sup$coord" 
## 15 "$quali.sup$cos2"  
## 16 "$quali.sup$v.test"
## 17 "$call"            
## 18 "$call$marge.col"  
## 19 "$call$marge.li"   
##    description                                           
## 1  "eigenvalues"                                         
## 2  "results for the variables"                           
## 3  "coord. of the categories"                            
## 4  "cos2 for the categories"                             
## 5  "contributions of the categories"                     
## 6  "v-test for the categories"                           
## 7  "results for the individuals"                         
## 8  "coord. for the individuals"                          
## 9  "cos2 for the individuals"                            
## 10 "contributions of the individuals"                    
## 11 "results for the supplementary quantitative variables"
## 12 "coord. of the supplementary quantitative variables"  
## 13 "results for the supplementary categorical variables" 
## 14 "coord. for the supplementary categories"             
## 15 "cos2 for the supplementary categories"               
## 16 "v-test for the supplementary categories"             
## 17 "intermediate results"                                
## 18 "weights of columns"                                  
## 19 "weights of rows"

MCA summary

For the variables a correlation ratio (squared) between it and each dimension is given (eta^2) enabling the plotting of the variables. The v-test in the summary follows a gaussian distribution referring to the category having a coordinate significantly different from zero.

summary(res_mca,abbrev=TRUE)
## 
## Call:
## MCA(X = dfalc, quanti.sup = (38:39), quali.sup = (32:37), graph = FALSE) 
## 
## 
## Eigenvalues
##                        Dim.1   Dim.2   Dim.3   Dim.4   Dim.5   Dim.6
## Variance               0.291   0.183   0.124   0.098   0.087   0.075
## % of var.             18.780  11.818   8.027   6.300   5.630   4.862
## Cumulative % of var.  18.780  30.597  38.625  44.925  50.555  55.417
##                        Dim.7   Dim.8   Dim.9  Dim.10  Dim.11  Dim.12
## Variance               0.068   0.058   0.051   0.047   0.043   0.040
## % of var.              4.368   3.777   3.269   3.023   2.765   2.603
## Cumulative % of var.  59.786  63.563  66.832  69.855  72.620  75.224
##                       Dim.13  Dim.14  Dim.15  Dim.16  Dim.17  Dim.18
## Variance               0.039   0.036   0.033   0.032   0.027   0.025
## % of var.              2.495   2.295   2.155   2.044   1.745   1.615
## Cumulative % of var.  77.719  80.014  82.168  84.212  85.957  87.572
##                       Dim.19  Dim.20  Dim.21  Dim.22  Dim.23  Dim.24
## Variance               0.025   0.023   0.020   0.018   0.015   0.012
## % of var.              1.606   1.507   1.289   1.145   0.982   0.778
## Cumulative % of var.  89.178  90.685  91.974  93.118  94.100  94.878
##                       Dim.25  Dim.26  Dim.27  Dim.28  Dim.29  Dim.30
## Variance               0.011   0.010   0.009   0.008   0.007   0.006
## % of var.              0.739   0.636   0.575   0.523   0.467   0.370
## Cumulative % of var.  95.617  96.253  96.829  97.351  97.818  98.188
##                       Dim.31  Dim.32  Dim.33  Dim.34  Dim.35  Dim.36
## Variance               0.005   0.005   0.004   0.003   0.003   0.002
## % of var.              0.346   0.319   0.251   0.221   0.200   0.142
## Cumulative % of var.  98.534  98.853  99.104  99.325  99.524  99.666
##                       Dim.37  Dim.38  Dim.39  Dim.40  Dim.41  Dim.42
## Variance               0.002   0.001   0.001   0.001   0.000   0.000
## % of var.              0.121   0.084   0.053   0.046   0.017   0.012
## Cumulative % of var.  99.787  99.872  99.925  99.971  99.988 100.000
## 
## Individuals (the 10 first)
##                            Dim.1    ctr   cos2    Dim.2    ctr   cos2  
## 1                       | -0.613  3.006  0.233 |  0.383  1.863  0.091 |
## 2                       |  0.638  3.253  0.121 | -0.088  0.098  0.002 |
## 3                       |  0.718  4.121  0.288 |  0.203  0.524  0.023 |
## 4                       | -0.399  1.274  0.162 | -0.321  1.307  0.105 |
## 5                       | -0.703  3.957  0.363 |  0.304  1.171  0.068 |
## 6                       |  0.837  5.599  0.303 |  0.399  2.027  0.069 |
## 7                       | -0.366  1.069  0.112 |  0.246  0.771  0.051 |
## 8                       |  0.156  0.196  0.014 | -0.169  0.364  0.016 |
## 9                       |  0.135  0.146  0.017 | -0.610  4.724  0.349 |
## 10                      | -0.257  0.529  0.067 |  0.186  0.439  0.035 |
##                          Dim.3    ctr   cos2  
## 1                        0.484  4.392  0.146 |
## 2                        0.205  0.783  0.012 |
## 3                       -0.130  0.317  0.009 |
## 4                       -0.318  1.890  0.103 |
## 5                        0.278  1.451  0.057 |
## 6                       -0.220  0.909  0.021 |
## 7                       -0.346  2.239  0.100 |
## 8                       -0.717  9.624  0.293 |
## 9                       -0.103  0.197  0.010 |
## 10                      -0.473  4.191  0.228 |
## 
## Categories (the 10 first)
##                            Dim.1    ctr   cos2 v.test    Dim.2    ctr
## (-Inf,57]               |  1.309  5.309  0.664  5.280 | -0.223  0.244
## (147, Inf]              | -0.923  2.416  0.293 -3.506 | -0.123  0.068
## (57,147]                | -0.278  0.399  0.067 -1.681 |  0.201  0.332
## (-Inf,49]               |  0.951  2.564  0.311  3.612 |  0.416  0.779
## (151, Inf]              | -0.425  0.513  0.062 -1.615 |  0.160  0.116
## (49,151]                | -0.275  0.411  0.072 -1.743 | -0.302  0.784
## B_MG_R_FR_allinallout_0 |  0.444  1.274  0.274  3.394 | -0.166  0.283
## B_MG_R_FR_allinallout_1 | -0.617  1.769  0.274 -3.394 |  0.231  0.393
## B_pestcontrolplan_no    |  0.195  0.354  0.196  2.869 | -0.085  0.108
## B_pestcontrolplan_yes   | -1.004  1.821  0.196 -2.869 |  0.439  0.554
##                           cos2 v.test    Dim.3    ctr   cos2 v.test  
## (-Inf,57]                0.019 -0.899 |  0.459  1.525  0.081  1.850 |
## (147, Inf]               0.005 -0.467 |  0.695  3.207  0.166  2.641 |
## (57,147]                 0.035  1.217 | -0.658  5.219  0.376 -3.974 |
## (-Inf,49]                0.059  1.579 |  0.410  1.114  0.058  1.557 |
## (151, Inf]               0.009  0.609 |  0.712  3.369  0.174  2.707 |
## (49,151]                 0.087 -1.910 | -0.588  4.378  0.330 -3.721 |
## B_MG_R_FR_allinallout_0  0.038 -1.269 | -0.195  0.571  0.053 -1.486 |
## B_MG_R_FR_allinallout_1  0.038  1.269 |  0.270  0.794  0.053  1.486 |
## B_pestcontrolplan_no     0.038 -1.256 | -0.148  0.476  0.113 -2.175 |
## B_pestcontrolplan_yes    0.038  1.256 |  0.761  2.448  0.113  2.175 |
## 
## Categorical variables (eta2)
##                                     Dim.1 Dim.2 Dim.3  
## MG_SOWS_perworkeredit_57_113_147_ | 0.732 0.037 0.383 |
## MG_R_PR_sowsinsection             | 0.314 0.095 0.341 |
## B_MG_R_FR_allinallout             | 0.274 0.038 0.053 |
## B_pestcontrolplan                 | 0.196 0.038 0.113 |
## BASIC_size_sows                   | 0.646 0.053 0.334 |
## M_farNSAIDS100                    | 0.096 0.405 0.028 |
## M_pregAB100                       | 0.037 0.243 0.022 |
## M_rOX                             | 0.212 0.206 0.007 |
## M_farAB100                        | 0.078 0.307 0.002 |
## M_pregNSAIDS100                   | 0.009 0.375 0.007 |
## 
## Supplementary categories (the 10 first)
##                            Dim.1   cos2 v.test    Dim.2   cos2 v.test  
## OUT_SOW_mort_dic_0      |  0.380  0.138  2.409 |  0.300  0.086  1.900 |
## OUT_SOW_mort_dic_1      | -0.363  0.138 -2.409 | -0.286  0.086 -1.900 |
## OUT_SOW_cull_dic_0      |  0.273  0.078  1.808 | -0.155  0.025 -1.029 |
## OUT_SOW_cull_dic_1      | -0.286  0.078 -1.808 |  0.163  0.025  1.029 |
## OUT_mort15_0            |  0.153  0.103  2.076 | -0.041  0.007 -0.552 |
## OUT_mort15_1            | -0.670  0.103 -2.076 |  0.178  0.007  0.552 |
## OUT_mort5_0             |  0.326  0.046  1.389 |  0.244  0.026  1.043 |
## OUT_mort5_1             | -0.141  0.046 -1.389 | -0.106  0.026 -1.043 |
## OUT_cull50_0            |  0.058  0.025  1.031 |  0.020  0.003  0.363 |
## OUT_cull50_1            | -0.438  0.025 -1.031 | -0.154  0.003 -0.363 |
##                          Dim.3   cos2 v.test  
## OUT_SOW_mort_dic_0      -0.113  0.012 -0.714 |
## OUT_SOW_mort_dic_1       0.108  0.012  0.714 |
## OUT_SOW_cull_dic_0      -0.077  0.006 -0.513 |
## OUT_SOW_cull_dic_1       0.081  0.006  0.513 |
## OUT_mort15_0            -0.101  0.044 -1.367 |
## OUT_mort15_1             0.441  0.044  1.367 |
## OUT_mort5_0              0.001  0.000  0.005 |
## OUT_mort5_1             -0.001  0.000 -0.005 |
## OUT_cull50_0            -0.032  0.008 -0.578 |
## OUT_cull50_1             0.246  0.008  0.578 |
## 
## Supplementary categorical variables (eta2)
##                           Dim.1 Dim.2 Dim.3  
## OUT_SOW_mort_dic        | 0.138 0.086 0.012 |
## OUT_SOW_cull_dic        | 0.078 0.025 0.006 |
## OUT_mort15              | 0.103 0.007 0.044 |
## OUT_mort5               | 0.046 0.026 0.000 |
## OUT_cull50              | 0.025 0.003 0.008 |
## OUT_cull30              | 0.122 0.008 0.023 |
## 
## Supplementary continuous variables
##                            Dim.1    Dim.2    Dim.3  
## OUT_SOW_mort_proNUM     | -0.371 | -0.148 |  0.143 |
## OUT_SOW_cull_proNUM     | -0.413 |  0.040 | -0.017 |

MCA dimensions

Dimdesc function points out the variables and the categories that are the most characteristic according to each dimension obtained by a MCA, i.e. it aims to facilitate interpretations of the dimensions in allowing to see which variables the axes are the most linked to/ which categories describe the best each axis.

dimdesc(res_mca,axes=1:2,proba=0.05)
## $`Dim 1`
## $`Dim 1`$quanti
##                     correlation p.value
## OUT_SOW_mort_proNUM       -0.37   0.014
## OUT_SOW_cull_proNUM       -0.41   0.006
## 
## $`Dim 1`$quali
##                                           R2         p.value
## MG_SOWS_perworkeredit_57_113_147_      0.732 0.0000000000036
## BASIC_size_sows                        0.646 0.0000000009650
## R_BR_floorsolid_0981_2                 0.560 0.0000000078485
## MG_BR_bedmatamount_no_alot_enough_some 0.551 0.0000001089746
## R_BR_sowspersection                    0.573 0.0000002462720
## R_BR_PREGsame                          0.473 0.0000003528065
## R_FAR_floorsolid_all0_100_100_2_muu1   0.508 0.0000007019593
## MG_FAR_bedamount                       0.471 0.0000029377317
## MG_FAR_nestmatamount                   0.436 0.0000104497923
## R_BR_kuivaliete                        0.379 0.0000110470059
## MG_BR_feedtype                         0.434 0.0000114838205
## MG_FAR_rootamount                      0.386 0.0000588893753
## MG_PR_bedmatamount_no_alot_enough_some 0.378 0.0000749701985
## B_MG_R_FR_allinallout                  0.274 0.0003136829729
## MG_R_PR_sowsinsection                  0.314 0.0005270298196
## MG_BR_animdirtmed                      0.242 0.0008170902172
## M_rOX                                  0.212 0.0018921444548
## MG_PR_kuivaliete                       0.212 0.0019003136639
## B_pestcontrolplan                      0.196 0.0029463577251
## R_PR_floorsolid_0791_2                 0.181 0.0044762285808
## MG_PR_rootamount_no_alot_some          0.207 0.0096180877367
## R_PR_dirtmed                           0.142 0.0128782579418
## OUT_SOW_mort_dic                       0.138 0.0141077568673
## OUT_cull30                             0.122 0.0215292476732
## OUT_mort15                             0.103 0.0362672362173
## MG_PR_animdirtmed                      0.102 0.0370065871473
## M_farNSAIDS100                         0.096 0.0434002412391
## 
## $`Dim 1`$category
##                                             Estimate       p.value
## (-Inf,57]                                      0.687 0.00000000003
## (-Inf,102]                                     0.664 0.00000000222
## R_BR_floorsolid_0981_2_2                       0.547 0.00000000785
## R_BR_PREGsame_1                                0.413 0.00000035281
## trough                                         0.393 0.00000915457
## MG_FAR_bedamount_1                             0.641 0.00000981448
## R_BR_kuivaliete_12                             0.362 0.00001104701
## R_FAR_floorsolid_all0_100_100_2_muu1_2         0.794 0.00004441590
## (-Inf,49]                                      0.468 0.00010357028
## MG_BR_bedmatamount_no_alot_enough_some_ALOT    0.583 0.00010450475
## all                                            0.441 0.00027637128
## MG_FAR_rootamount_1                            0.682 0.00028443403
## B_MG_R_FR_allinallout_0                        0.286 0.00031368297
## MG_BR_animdirtmed_1                            0.269 0.00081709022
## M_rOX_0                                        0.254 0.00189214445
## MG_PR_kuivaliete_12                            0.254 0.00190031366
## B_pestcontrolplan_no                           0.323 0.00294635773
## R_PR_floorsolid_0791_2_2                       0.234 0.00447622858
## <20                                            0.489 0.00473926869
## MG_FAR_nestmatamount_1                         0.522 0.01133726527
## R_PR_dirtmed_1                                 0.204 0.01287825794
## OUT_SOW_mort_dic_0                             0.201 0.01410775687
## MG_FAR_nestmatamount_2                         0.053 0.01618214072
## OUT_cull30_0                                   0.201 0.02152924767
## MG_PR_bedmatamount_no_alot_enough_some_ALOT    0.303 0.02153413745
## MG_FAR_ind_0no_1rout_2sometimes_0              0.281 0.03257320959
## MG_PR_rootamount_no_alot_some_ALOT             0.319 0.03593801699
## OUT_mort15_0                                   0.222 0.03626723622
## MG_PR_animdirtmed_1                            0.181 0.03700658715
## M_farNSAIDS100_1                               0.168 0.04340024124
## M_farNSAIDS100_2                              -0.168 0.04340024124
## MG_PR_animdirtmed_2                           -0.181 0.03700658715
## OUT_mort15_1                                  -0.222 0.03626723622
## OUT_cull30_1                                  -0.201 0.02152924767
## OUT_SOW_mort_dic_1                            -0.201 0.01410775687
## R_PR_dirtmed_2                                -0.204 0.01287825794
## MG_FAR_rootamount_0                           -0.621 0.00752282620
## MG_PR_rootamount_no_alot_some_0               -0.410 0.00595920041
## R_PR_floorsolid_0791_2_1                      -0.234 0.00447622858
## B_pestcontrolplan_yes                         -0.323 0.00294635773
## MG_PR_kuivaliete_2                            -0.254 0.00190031366
## M_rOX_1                                       -0.254 0.00189214445
## MG_BR_animdirtmed_2                           -0.269 0.00081709022
## (635, Inf]                                    -0.493 0.00079474026
## R_FAR_floorsolid_all0_100_100_2_muu1_0        -0.657 0.00041785500
## B_MG_R_FR_allinallout_1                       -0.286 0.00031368297
## (147, Inf]                                    -0.517 0.00018011497
## MG_FAR_bedamount_0                            -0.508 0.00011727921
## 50-100                                        -0.631 0.00002175542
## MG_PR_bedmatamount_no_alot_enough_some_0      -0.481 0.00001571214
## MG_FAR_nestmatamount_0                        -0.574 0.00001231460
## R_BR_kuivaliete_2                             -0.362 0.00001104701
## MG_BR_bedmatamount_no_alot_enough_some_no     -0.577 0.00000219720
## crate_L                                       -0.594 0.00000173634
## R_BR_PREGsame_0                               -0.413 0.00000035281
## R_BR_floorsolid_0981_2_1                      -0.547 0.00000000785
## 
## 
## $`Dim 2`
## $`Dim 2`$quali
##                                          R2     p.value
## MG_PR_bedmatamount_no_alot_enough_some 0.58 0.000000037
## M_farNSAIDS100                         0.41 0.000004478
## M_pregNSAIDS100                        0.37 0.000012953
## MG_PR_rootamount_no_alot_some          0.40 0.000034516
## M_farAB100                             0.31 0.000114170
## MG_PR_kuivaliete                       0.29 0.000210855
## MG_FAR_dirtmed                         0.27 0.000348390
## R_PR_dirtmed                           0.25 0.000736204
## M_pregAB100                            0.24 0.000786324
## R_PR_floorsolid_0791_2                 0.22 0.001674460
## M_rOX                                  0.21 0.002219612
## MG_FAR_bedamount                       0.25 0.002884813
## MG_BR_bedmatamount_no_alot_enough_some 0.24 0.003697582
## MG_FAR_ind_0no_1rout_2sometimes        0.24 0.003741257
## MG_FAR_ox                              0.23 0.004800378
## R_FAR_floorsolid_all0_100_100_2_muu1   0.23 0.005722594
## R_BR_kuivaliete                        0.13 0.016883835
## 
## $`Dim 2`$category
##                                             Estimate    p.value
## MG_PR_bedmatamount_no_alot_enough_some_ALOT    0.402 0.00000075
## M_farNSAIDS100_2                               0.274 0.00000448
## M_pregNSAIDS100_2                              0.285 0.00001295
## MG_PR_rootamount_no_alot_some_ALOT             0.325 0.00001465
## M_farAB100_2                                   0.239 0.00011417
## MG_PR_kuivaliete_12                            0.235 0.00021086
## MG_FAR_dirtmed_1                               0.224 0.00034839
## R_PR_dirtmed_1                                 0.213 0.00073620
## M_pregAB100_2                                  0.221 0.00078632
## MG_BR_bedmatamount_no_alot_enough_some_ALOT    0.401 0.00090371
## R_PR_floorsolid_0791_2_2                       0.203 0.00167446
## M_rOX_1                                        0.199 0.00221961
## MG_FAR_bedamount_0                             0.193 0.00533723
## R_FAR_floorsolid_all0_100_100_2_muu1_0         0.176 0.00641463
## R_BR_kuivaliete_12                             0.169 0.01688384
## MG_FAR_ind_0no_1rout_2sometimes_2              0.089 0.02093014
## MG_FAR_rootamount_0                            0.244 0.04132871
## (-Inf,3]                                       0.118 0.04887658
## R_BR_kuivaliete_2                             -0.169 0.01688384
## M_rOX_0                                       -0.199 0.00221961
## R_PR_floorsolid_0791_2_1                      -0.203 0.00167446
## R_FAR_floorsolid_all0_100_100_2_muu1_1        -0.271 0.00130594
## (3,7]                                         -0.285 0.00105264
## MG_FAR_ind_0no_1rout_2sometimes_0             -0.315 0.00096800
## M_pregAB100_1                                 -0.221 0.00078632
## R_PR_dirtmed_2                                -0.213 0.00073620
## MG_FAR_bedamount_2                            -0.309 0.00063319
## MG_FAR_dirtmed_2                              -0.224 0.00034839
## MG_PR_kuivaliete_2                            -0.235 0.00021086
## M_farAB100_1                                  -0.239 0.00011417
## MG_PR_rootamount_no_alot_some_HIE             -0.270 0.00009645
## M_pregNSAIDS100_1                             -0.285 0.00001295
## M_farNSAIDS100_1                              -0.274 0.00000448
## MG_PR_bedmatamount_no_alot_enough_some_NIU    -0.381 0.00000201

Eigenvalues and scree plot

The proportion of variances retained by the different dimensions (axes) can be extracted separately.

eig.val <- get_eigenvalue(res_mca)
head(eig.val,n=10)
##        eigenvalue variance.percent cumulative.variance.percent
## Dim.1       0.291             18.8                          19
## Dim.2       0.183             11.8                          31
## Dim.3       0.124              8.0                          39
## Dim.4       0.098              6.3                          45
## Dim.5       0.087              5.6                          51
## Dim.6       0.075              4.9                          55
## Dim.7       0.068              4.4                          60
## Dim.8       0.058              3.8                          64
## Dim.9       0.051              3.3                          67
## Dim.10      0.047              3.0                          70

To visualize the percentage of inertia explained by each MCA dimension:

eig.val <- res_mca$eig
barplot(eig.val[, 2], 
        names.arg = 1:nrow(eig.val), 
        main = "Variances Explained by Dimensions (%)",
        xlab = "Principal Dimensions",
        ylab = "Percentage of variances",
        col ="steelblue")
# Add connected line segments to the plot
lines(x = 1:nrow(eig.val), eig.val[, 2], 
      type = "b", pch = 19, col = "red")

Eigenvalues can be used to determine the number of axes to retain. As to my knowledge, there is no “rule of thumb” to choose the number of dimensions to keep for the data interpretation. It depends on the research question and the researcher’s need. The level of satisfaction, e.g.in case of 80% of the total variance explained the number of dimensions necessary to achieve that can be chosen.

The first two express 14% of the total dataset variance meaning that 14% of the individuals or variables total variability is explained by the plane. This is a very small percentage.In addition, there is no clear drop to help to identify how many dimensions should be included in the final interpretation to capture the right number of real information. Dimensions having low scores are likely to be unstable, too.

Graphical representation of individuals and variable categories

To further clarify the MCA results graphical representation is used. Firstly, a biplot showing the global pattern within the data is created. Observations are represented by blue points and variables by red triangles and labels. The distance between any observation points or variable points gives a measure of their similarity (or dissimilarity). Similar types of individuals are close on the map, as well as similar kinds of variables.

fviz_mca_biplot(res_mca, 
               repel = TRUE, # Avoid text overlapping (slow if many point)
               ggtheme = theme_minimal())

Graphical representation of variables

Variable categories related results can be extracted separately to provide information for the coordinates, the cos2 and the contribution of variable categories:

  • var$coord: coordinates of variables to create a scatter plot

  • var$cos2: represents the quality of the representation for variables on the factor map.

  • var$contrib: contains the contributions (in percentage) of the variables to the definition of the dimensions.

Next, a plot is created to visualize the correlation between variables of the first and second dimension. Basically variable categories with a similar profile are grouped together. Negatively correlated variable categories are positioned on opposite sides of the plot origin (opposed quadrants). The distance between category points and the origin measures the quality of the variable category on the factor map. Category points that are away from the origin are well represented on the factor map. Supplementary quantitative final grade variable is plotted blue.

fviz_mca_var(res_mca, choice = "mca.cor", 
            repel = TRUE, # Avoid text overlapping (slow)
            ggtheme = theme_minimal())

The plot should help to identify variables that are the most correlated with each dimension. The squared correlations between variables and the dimensions are used as coordinates.

It can be seen that, the variables mother´s and father´s education as well as father´s job are the most correlated with dimension 1. Similarly, the variables going out with friends, high alcohol usage and class failures are the most correlated with dimension 2.

Variable contribution

It’s possible to change the color and the shape of the variable points as well as the number of top variables, i.e. the ones having the highest contribution.

fviz_mca_var(res_mca, col.var="black", shape.var = 15,
             repel = TRUE,select.var = list(contrib = 6))

It is also possible to control the transparency of variable categories according to their contribution values.

# Change the transparency by contrib values
fviz_mca_var(res_mca, alpha.var="contrib",
             repel = TRUE,
             ggtheme = theme_minimal())

The most contributing, i.e. important variable categories can be visualized by gradient-colouring them respect to their contribution value. Meaning basically, that low, medium and high contributions have different colours.

fviz_mca_var(res_mca, col.var = "contrib",
             gradient.cols = c("#00AFBB", "#E7B800", "#FC4E07"), 
             repel = TRUE, # avoid text overlapping (slow)
             ggtheme = theme_minimal()
             )

Simple bar plots can also be used to visualize contribution of variable categories. The top 12 variable categories contributing to the first and second dimension:

# Contributions of rows to dimension 1
fviz_contrib(res_mca, choice = "var", axes = 1, top = 12)

# Contributions of rows to dimension 2
fviz_contrib(res_mca, choice = "var", axes = 2, top = 12)

The red dashed line indicates the expected average value, If the contributions were uniform.

Variable cos2

For cos2 values it is similarly possible to change the color and the shape of the variable points as well as the number of top variables, i.e. the ones having the highest cos2 values.

fviz_mca_var(res_mca, col.var="black", shape.var = 15,
             repel = TRUE,select.var = list(cos2 = 10))

If a variable category is well represented by two dimensions, the sum of the cos2 is closed to one. For some of the row items, more than 2 dimensions are required to perfectly represent the data. Or it’s transparency of the variable categories according to their cos2 values can be controlled.

# Change the transparency by cos2 values
fviz_mca_var(res_mca, alpha.var="cos2",
             repel = TRUE,
             ggtheme = theme_minimal(),selectMod="cos2 10")

Furthermore, just as with the contributions, variable categories can be gradient-coloured with respect to their cos2 value. Meaning that low, medium and high co2 values have different colours.

# Color by cos2 values: quality on the factor map
fviz_mca_var(res_mca, col.var = "cos2",
             gradient.cols = c("#00AFBB", "#E7B800", "#FC4E07"), 
             repel = TRUE, # Avoid text overlapping
             ggtheme = theme_minimal())

Similarly as with the contributions, it is also possible to create a bar plot of variable cos2 importance.

# Cos2 of variable categories on Dim.1 and Dim.2
fviz_cos2(res_mca, n=10,choice = "var", axes = 1:2,top=12)

Grouped biplots

Individuals can be coloured by groups and a concentration ellipse can be added around each group.

Grouped biplots:Performance group

# habillage = external grouping variable
fviz_mca_ind(res_mca, habillage = dfalc$OUT_SOW_mort_dic, addEllipses = TRUE)

Grouped biplots:High alcohol

# habillage = index of the column to be used as grouping variable
fviz_mca_ind(res_mca, habillage = dfalc$OUT_SOW_cull_dic, addEllipses = TRUE)

Grouped biplots:Performance group and high alcohol

fviz_ellipses(res_mca, c("OUT_SOW_cull_dic", "OUT_SOW_mort_dic"),
              geom = "point")

Quantitative supplementary variable:Final grade

fviz_mca_var(res_mca, choice = "quanti.sup",
             ggtheme = theme_minimal())

res_mca$quanti
## $coord
##                     Dim 1 Dim 2  Dim 3  Dim 4 Dim 5
## OUT_SOW_mort_proNUM -0.37 -0.15  0.143 -0.203 -0.10
## OUT_SOW_cull_proNUM -0.41  0.04 -0.017  0.047  0.12

Summary of the first plane

First dimension

The first dimension aims to characterize individuals with a high positive coordinate on the axis (right) and with a high negative coordinate (left).

Firstly, there are a group of individuals on the right with both mother and father having higher education, mother and father working as a teacher, no class failures and the best performance group. Additionally, they are sharing being active, young, having a mother work in health sector and having family support. On the contrary, there are low frequency scores for mother and father having only elementary level or no education, mother or father work as “other”" or mother being at home and having class failures as well as low frequencies for no activities, belonging to the lowest performance group category, average health group and no family support.

Secondly, there are a group of individuals on the left with class failures one or more, lowest performancegroup, no willingness to higher education, high alcohol consumption and the oldest age category. Additionally, to some extent common are mother having low educational level, going out with friends, mother work as other or have average level education as well as being a male. Low frecuencies are for no class failures, mother and father being high school educated, young age group, mother working as a teacher, best performance group, being a female and having father as a teacher.

Thirdly, there is a group on the left with mother and father having little education, health being average, gender female, low alcohol consumption, mother working at home and father as “other”“, having a positive attitude towards high edution, performing at an almost average level and sometimes going out with friends. Low frequencies are there for mother´s or father´s higher education, mother working as a teacher, being a male, using alcohol, going out a lot, father being a teacher, health being very good, having no ideas about further education and haing some class failures.

Second dimension

The second dimension aims to characterize individuals with a high positive coordinate on the axis (top) and with a high negative coordinate (bottom).

Firstly, there is a group up in the graph having in common one or more class failures, lowest performance group, negative attitude towards higher eduction, high alcohol consumption, age above 17, father´s low education, going out frequently, mother working as “other”, mother´s low education and being a male. Additionally, the group members have low frequencies for no class failures, mother´s and father´s high education, positive attitude towards education, low alcohol consumption, being young, having mother in education, best performance group, being a female and having father in education.

Secondly, there is a group sharing high frequency for mother and father being highly educated, mother and father working as educators, no class failures, performing best, being active and young, mother working in the health sector having family support. In addition there is low frequency for mother´s or father´s low education, jobs as “other”“, mother at home, one or more class failures, no activities, lowest performance group, average health and no family support.

Thirdly, there is a negative co-ordinate group sharing high frequency for the lowest level of education of the mother and the father, average health, being a female, low alcohol usage, mother being at home, father working as “other”, performing middle low, rarely or never going out and having a positive attitude towards education. Low frequencies are common for high education for mother and father, mother being a teacher, being a male, using a lot of alcohol, going out frequently, father working as a teacher, best health group, no positive attitude towards education, one or more failures.

Clusters

Finally, a classification made on individuals reveals three clusters.

The first cluster has individuals with high frequencies for

and low frequencies for

The second cluster has individuals with high frequencies for

and low frequencies for

And, finally, the third cluster has individuals with high frequencies for

and low frequencies for

Conclusion

The aim of this study was to scrutinize the multidimensional data into a more comprehensible, lower dimensional structure and to reveal some association between different types of respondents. However, the reduction was not completely suitably acchieved as the inertia explained was low and the observations were very scattered. However, there is an indication that some sociodemographic factors have joint effects. It is important to confirm the associations using advanced techniques, e.g. by applying theory of planned behavior to study the relations among personal beliefs, attitudes, behavioral intentions and behaviour and other individual as well as parental features to investigate the risk factors for high alcohol usage. Future investigations need to be done to identify those variables that show significant relationships and to take them forward for further analysis.

References: